# What is science?

We can read the steps of the scientific method and know about conventions such as the peer review process but at its core what is science? I have been thinking about this question more and more recently. In trying to pin it down the definition seems to diffuse into a overly abstract statement that resembles something I dislike as a shortcut to thinking and refer to as "bumper sticker philosophy." It seems odd to phrase it this way but science is, in a way, about what works and what doesn't work, and understanding the reasons why what works and what doesn't work in a very rigorous way. Science, ideally, is a careful balance of keeping an open mind, embracing curiosity, taking details seriously, and reasoned agreement and disagreement with others. Then there is the real world human aspect of how science is or is not practiced. Statements such as "believing" in evolution or "believing" in climate change do not sit well in a scientific perspective. In current American English this implies faith in something... However, pointing this out can be misinterpreted and hijacked by people with non-scientific motivations.

At the risk of sounding like a kōan, I thought two famous historical case examples, one from physics and one from biology, that have interested me for different reasons, might provide more of an illustration of what science is then trying to nail down a precise verbal/text definition. However, this comes at the risk of people not being patient enough to read them. Perhaps therein lies part of the modern problem. If it doesn't fit into a snappy short phrase it never reaches most people. I've linked PDFs of the two examples below.

# Article about misconceptions and science as a great equalizer

Right or wrong, I have avoided posting much here about politics; however, I just came across this article that refers to the upcoming March for Science and wanted to share it.

"Let science be science again" by Y. Cheng (http://thebulletin.org/let-science-be-science-again10668)

On a personal note, I (Floyd Reed) am from the rural southern Appalachian mountains. Just to put it bluntly, many of my family and friends that I grew up with are anti-intellectual. Many of these people would now consider me an "out of touch elite" because I am a scientist and a professor at a university while they are the disenfranchised "left-behind working class" to place it into some of the terms used in part of the current political dialogue. In some ways I feel like I have a fairly unique perspective in understanding some of the motivations leading to this. It is not easy to forget, for example, "inbreeding" jokes or condescending comments about my accent as a student within academia and to react by constructing an us-versus-them social polarization, which is then all-to-easily hijacked by politicians and others to further their interests. However, my connection to the same cultural background is also conveniently ignored by the people I grew up with---I am now an "outsider"---because it does not fit into the (overly polarized) social myth being constructed. Coming from this perspective points like this in the article stood out to me.

"Interestingly, both of these criticisms reflect a common misconception: that science is the elite pursuit of the privileged few. ... These folks, mostly men, fiddle on futuristic gadgets in secluded rooms inside ivory towers, high above the dwellings of earthly mortals who are not worth the scientists’ time or attention. However, I know that if I were a white man born and raised in America, I might not have become a scientist. The opportunities and resources America would have offered me might very well have led me to pursue another career path based on my broad interests as a child; ... but, fortunately, they were not prerequisites for a career in science. Science is a great equalizer, not a privilege. I believe in the promise of science because I have lived it. I’ll be marching as a personal testament to the promise, and to help keep it alive."

"Science is a great equalizer because it not only empowers the individual, but also lifts whole communities. It is often less-developed regions that more acutely suffer from the lack of access to science—and more significantly benefit from science’s equalizing power."

I also want to comment on where anti-science views fall within politics and who controls how this is framed. Many of the people reading this may view anti-science sentiments as corresponding with conservative political ideologies. For example, for whatever reasons anti-climate-change has been assigned to, or adopted by, the political right. But it can also be pointed out that if we are defending science from politics the current "liberal/left" also has its share of anti-scientific leanings such as aspects of the GMO crop debate. It is tempting to assign labels to an issue that corresponds to a political axis. Defending science from politics also means not playing into political labeling. Here is another quote from Cheng's article:

"Nature has no political ideology, nor should the interpretation of nature. Traditionally in the United States, basic research has enjoyed bipartisan support because science is nonpartisan. Anti-science views exist on both ends of the political spectrum. ... There have always been opponents of science who try to smear or discredit facts that counter their interest-driven agenda. Supporters of science cannot let the other side frame the debate. Staying silent for fear of being misunderstood is not self-preservation; ... The battle lines are drawn between truth and falsehoods, regardless of rhetoric or ideology. ... When ignorance is touted as a virtue, and anti-intellectualism is encroaching on the very fabric of our society, the fight for the preservation and advancement of science calls for a unified effort from the scientific community, and from supporters of science of every color and creed."

Finally, the term "science" is being thrown around a lot. However, it may mean different things to different people, which can get tricky when assumptions are being made about mutual understanding of the term. It is worth taking some time to think about this. In a future post I want to make some comments about what science is and is not, at least from my perspective.

# An underdominance "trigger" gene drive system

This is one of those lack-of-time-and-funding free idea posts. I had this idea several years ago but have not been able to act on it with all the other priorities I have to deal with. So, I am throwing it out there in case someone who can wants to explore it. If someone has the time and resources this could be done fairly quickly to see if it might work.

Double stranded RNA triggers the RISC (RNA Induced Silencing Complex) pathway and the cell degrades single stranded RNA that matched the double stranded sequence.

Imagine inverting part of a gene so that the resulting mRNA had a reverse sequence for a stretch of nucleotides, at least 21 nucleotides, seven or eight codons, and preferably longer. This might disrupt the amino acid sequence but if chosen carefully, the right gene and position, the disruptions could be minimal with conservative amino acid changes. It also might be in a UTR without disrupting function too much (or even a "tag" extension to a UTR).  (I am not sure if targeting an mRNA intron sequence would be efficient enough knockdown because splicing may occur at a high enough rate to evade degradation...?)

A homozygote for the sequence inversion could be fine. However, a heterozygote with wildtype could trigger the RISC system if the mRNAs bound together.  This would knockdown expression of the corresponding gene. While it might be viable (which would help to engineer the system in the first place), it could have a very strong fitness effect.

If heterozygotes have a lower fitness than either homozygote you have underdominance, which can be utilized to transform a wild population.

Originally I was thinking along the lines of a more complex two-locus system using phiC31 integration sites for transformations; however, cassette exchange could be used to modify a single locus (e.g., Zhang, X., Koolhaas, W. H., & Schnorrer, F. (2014). A versatile two-step CRISPR-and RMCE-based strategy for efficient genome engineering in Drosophila. G3: Genes| Genomes| Genetics, 4(12), 2409-2418.).

# Hawaiʻi Senate Resolution on Mosquito Technology

There was a resolution submitted to the state senate entitled "Requesting that the University of Hawai‘i provide information to the legislature on possible techniques to eliminate mosquitos from Hawaii."

Some of us working on mosquitoes in Hawai'i found out about this at the last minute. I and others submitted written testimony as soon as we learned of this (but it was within 24 hours of the session to consider the resolution and may not have been provided) and contacted DLNR and the university office that submits testimony on behalf of the university. The last I heard was that the resolution passed with some amendments. The full text is copied below.

THE SENATE         TWENTY-NINTH LEGISLATURE, 2017         STATE OF HAWAII

S.C.R. NO. 160

SENATE CONCURRENT RESOLUTION

Requesting that the University of Hawai‘i provide information to the legislature on possible techniques to eliminate mosquitos from Hawaii.

WHEREAS, mosquitos are an alien species in Hawaii and were introduced to the Hawaiian archipelago in the late 1800s; and

WHEREAS, mosquitos are vectors of diseases that affect humans and wildlife; and

WHEREAS, without sufficient eradication efforts, mosquito-borne diseases such as dengue fever and zika virus are likely to become established in Hawaii; and

WHEREAS, avian malaria has decimated the native bird population throughout the range of the mosquito; and

WHEREAS, as our climate warms, mosquitoes are able to survive at higher altitudes and Hawaii's native birds will continue to lose survivable habitat as mosquitos move mauka; and

WHEREAS, eliminating mosquitos is a critical element of restoring the original habitat of Hawaii's native birds and could be a deciding factor in saving native birds from extinction; and

WHEREAS, mosquito eradication techniques, including introducing genetically-engineered mosquitoes to disrupt breeding and reproduction cycles, have been developed and tested in multiple mosquito-affected regions of the world; now, therefore,

BE IT RESOLVED by the Senate of the Twenty-ninth Legislature of the State of Hawaii, Regular Session of 2017, the House of Representatives concurring, that the University of Hawai‘i, through its College of Tropical Agriculture and Human Resources and College of Agriculture, Forestry, and Natural Resource Management, is requested to investigate techniques that can be used to eliminate mosquitoes from Hawaii, the estimated cost of establishing a statewide mosquito eradication program using identified techniques, and the expected environmental impacts of such a program; and

BE IT FURTHER RESOLVED that the University of Hawai‘i is requested to submit a report of its findings and recommendations, including any proposed legislation, to the Legislature no later than twenty days prior to the convening of the Regular Session of 2018; and

BE IT FURTHER RESOLVED that certified copies of this Concurrent Resolution be transmitted to the Governor; Chancellor of the University of Hawai‘i System; Dean of the College of Tropical Agriculture and Human Resources; Dean of the College of Agriculture, Forestry, and Natural Resource Management; and Deputy Director of the Department of Health's Environmental Health Administration.

# Evolutionary Genetic Engineering in the Indo-Pacific: Conservation, Humanitarian, and Social Issues

Preface
The text in this post is from a "perspective" manuscript I was asked to write last year for a special issue of a journal focused on gene drive technologies. Long story short this manuscript was rejected and I see little recourse to submit it elsewhere for publication in a timely manner. However, I feel like many of the points in the manuscript should be made, at least in some form; so, I am posting it here.

Abstract
The Indo-Pacific region contains a unique mix of opportunities for the development and use of genetic-pest-management, gene-drive, and gene-drive-like technologies. Here I collectively refer to these technologies as Evolutionary Genetic Engineering (EGE). Indo-Pacific Islands have some of the world's highest rates of endemism and extinction—species and entire ecosystems are at risk. This threat to the natural world is coupled with the burden of human diseases, many of which are new and emerging or neglected tropical diseases. The same factors which have led to high rates of endemism also, in some ways, make this region an ideal testing ground for some types of EGE's. There is great potential for positive humanitarian, economic, and conservation applications of EGE's. However, these types of new technologies will be initially viewed from the perspective of the recent history of a loss of self determination, issues of social justice, and the testing of new technologies (e.g., biocontrol, agricultural, nuclear) in the Indo-Pacific—a region of the world that is still extensively colonized and controlled by Western Nations. Experience with successes and failures in related technologies suggests a path to move forward—a set of eight recommendations—to maximize the potential payoffs and minimize unintended negative effects of EGE's.

Introduction
The island Indo-Pacific is a large, important, unique, and unfortunately often overlooked region of the world. There is tremendous potential for the positive use of Evolutionary Genetic Engineering (EGE) in the region in both humanitarian and conservation applications. This potential stems from the regions geographic isolation, collection of infectious diseases, and species conservation urgencies. However, it would be a mistake to neglect the context of recent and ongoing political and social challenges in the region. Doing so is likely to generate a negative reaction that could inhibit the applications of promising emerging technologies. This context includes issues of colonialism, self determination, biocontrol, the testing of new technologies, and early experiences with genetically modified agricultural crops in the region. In this article I am focused on the island Indo-Pacific tropics, but also use examples from the broader region including India and Australia. I am also focusing on terrestrial applications of EGE’s. There are potential freshwater and marine applications, but this is less developed and goes beyond the scope of the current article.

In order to move forward in a way that does not sacrifice long term progress for short term convenience, we must accept that everyone has a role to play in shaping our technological future; this is not always easy to do when faced with confrontations and fundamental disagreements. To do this we must

1. enhance communications and avoid a reluctance to provide more detailed information about new technologies or to be dismissive of inquiries.
2. EGE applications should only be pursued if there is a genuine benefit to the local population (and if the people potentially affected generally agree that this is desirable rather than the decision being made externally), not in order to test new technologies in a “safe” manner or to avoid jurisdictional regulations.
3. The potential benefits and risks of EGE's, along with the degree of uncertainty surrounding both, need to be unambiguously communicated.
4. There needs to be a frank discussion of unintended side effects and the potential for misuse of the technology.
5. Humanitarian goals need to be administered and controlled by humanitarian organizations while conservation goals need to be administered and under the control of conservation organizations (applicable to both governmental and non-governmental organizations); this is especially true in an international setting.
6. Proactive research needs to be conducted and the data available to address common concerns about the possible ecological and health effects of EGE’s.
7. It takes broad perspectives, beyond what any single person is capable of, to identify potential promises and pitfalls of the development and implementations of EGE's.
8. Finally, a broad-based community discussion of, and direct involvement in, EGE development and applications should occur as early as possible. This will positively shape both the development and applications of the technology and help build a solid social foundation for future developments.

Potential Evolutionary Genetic Engineering Applications in the Island Indo-Pacific.
The Indo-Pacific spans half of the Earth’s circumference yet receives relatively less international focus. A revealing example is that this is the non-polar region that is most often divided on map projections of the world. This inattention is not simply due to a smaller population; the four most populous countries in the world (China, India, the US, and Indonesia) have territory and active interests in the region.

The island Indo-Pacific has some of the world’s highest rates of both species endemism (unique genetic diversity) and extinction (Vitousek 1988; Fleischer et al. 1998; Myers et al. 2000; Kier et al. 2009). Adaptive radiations of species here have served as prime examples of evolutionary biology (p. 380 Darwin 1845; Dobzhansky 1973). Extinctions in these species-rich regions are proceeding at an alarming rate (Pimm et al. 1995; Ganzhorn et al. 2001; Fonseca 2009; Loehle & Eschenbach 2012; Régnier et al. 2015) and this is predicted to be exacerbated by climate change (Benning et al. 2002; Mora et al. 2013). The region is in dire need of effective conservation strategies and potential Evolutionary Genetic Engineering (EGE) applications targeting introduced species and diseases have been proposed to establish effector genes refractory to introduced vectored disease (genetic modifications to block transmission of the disease) and genetic sterile insect techniques to suppress populations of invasive species (Clarke 2002; Wimmer 2005; Altrock et al. 2010; Esvelt et al. 2014; Reeves et al. 2014; Webber et al. 2015; see Sinkins and Gould 2006 and Gould et al. 2006 for review).

The island Indo-Pacific is also home to newly emerging and/or neglected tropical diseases that affect human health as well as economically important species. Vector borne human diseases in the region include chikungunya, dengue fever, Japanese encephalitis, lymphatic filariasis, malaria, plague and Rift Valley fever (Madagascar), schistosomiasis, scrub typhus, West Nile fever, and zika. Additionally there are diverse agricultural crop pests and diseases that impact food production across the region. A major goal of EGE development is to address human disease, and there is also potential for agricultural applications (Alphey 2002; Sinkins and Gould 2006; Gould 2008; Wimmer 2013; Esvelt et al. 2014; Champer et al. 2016).

When countries are listed by gross domestic product per capita it becomes apparent that the Indo-Pacific contains, in terms of national economic wealth, many of the poorest countries in the world. For example, the Comoros, Kiribati, Madagascar, Marshall Islands, Micronesia, Papua New Guinea, Solomon Islands, Tuvalu, and Vanuatu have an average per capita GDP of Intl.$2,387, approximately 1/8th of the world average, Intl.$18,872 (IMF 2016). This limits the resources available that these governments can apply to humanitarian and conservation interventions and suggests an enhanced value of international collaboration.

In many ways the terrestrial isolation that has led to the Indo-Pacific's tremendous biological diversity also makes the region ideal for some EGE applications. Suppressing or modifying non-native invasive species is an obvious place to start. However, what may be a pest in one location may be a highly valued or important ecological species in another (e.g., nopal Opuntia cacti are a highly valued component of Mexican cuisine and source of animal fodder while considered an invasive species pest in Australia—Cactoblastis cactorum has been used successfully as bio-control in Australia but is now threatening native Opuntia in the Americas, Zimmerman et al. 2004). Proper application of Type 1 and 3a EGE’s (Appendix A: Types of Evolutionary Genetic Engineering) can leave a species genetically unmodified within its native range, even with low levels of migration between islands or islands and continents (Altrock et al. 2010, 2011; Láruson & Reed 2016). The limited and discrete partitioning of land area of islands make 100% local genetic transformation or eradication of a species possible without resorting to type 3b or 4 EGE’s (Appendix A) and allows the application to proceed in a stepwise fashion across multiple islands using limited resources.

Colonialism, self determination, and the testing of new technologies
We have different perspectives depending on our experiences and social identities, and we are all-too-often not aware of how our individual perspective differs from others. In the middle of the abstract I used the following sentence, “The same factors which have led to high rates of endemism also, in some ways, make this region an ideal testing ground for some types of EGE's.” I chose the wording of this sentence carefully. What was your reaction? For many of the people reading this article the sentence seemed perfectly natural and flows into the ideas of the preceding and following sentences. However, for some readers the phrase “testing ground” is likely to stand out. Our reaction to this sentence is related to our perspective. For many who do not live in the island Indo-Pacific it is easy to see the region as something external to our daily lives and more disposable for testing and experimenting. In contrast, for some the Indo-Pacific represents home, family, work, and is also fundamentally connected to a cultural identity. I ask readers to construct your own sentence connecting a place that is highly valued to you personally (your hometown, where you live now, or a place of historical, religious, or cultural importance) with a “testing ground” of a new potentially powerful technology with its own set of concerns and unknowns.

The Indo-Pacific has a long and continuing history of a loss of self determination and sovereignty. The UN Special Committee on Decolonization lists American Samoa, French Polynesia, Guam, New Caledonia, Pitcairn, and Tokelau as Non-Self-Governing-Territories. The total number of ongoing sovereignty disputes encompasses many more islands and regions too extensive to list here. Colonization includes the establishment of extensive military bases and use of the islands for tests of nuclear, biological, and chemical warfare technologies—and these were not limited to a few isolated incidents, for example hundreds of nuclear weapons tests were conducted by France in Mururoa and Fangataufa Atolls, by the United Kingdom in South Australia, Montebello, and Kiritimati Islands, by the United States in Pikinni (Bikini), Ānewetak (Enewetak), Johnston (Kalama) Atolls, and Kiritimati. This history of military testing, non-military testing of new technologies (e.g., disastrous attempts at classical biological control by introducing new species, e.g., Howarth 1983; Clarke et al. 1984; Henneman & Memmott 2001; Messing & Wright 2006; Hays and Conant 2007; Parry 2009), and colonization in the region can severely inhibit international biological research and potential applications including EGE’s.

There is a case study that deserves special mention within the context of EGE’s in the Indo-Pacific, especially in the context of international programs and applications of mosquito genetic engineering. From 1969–1975 the World Health Organization (WHO) collaborated with the US Public Health Service (PHS) and the Indian Council of Medical Research (ICMR) to establish a Genetic Control of Mosquitoes Research Unit (GCMRU) in India; this was financially supported by the Government of India, US PL-480 funds, and the CDC (AEND 1975a). The GCMRU was studying and implementing mosquito control technologies including the release of sterilized individuals. What appears to have started with concerns about a carcinogen (thiotepa) being added to well water in the village of Pochanpur without public or government consultation got caught up in politics (AEND 1975b; Hanlon 1975), with widespread accusations in the media and later by the Government of India, and grew into a political disaster with suspicions that the US military was using India to test methods of biological warfare using mosquitoes (Sehgal 1974; Anonymous 1975; AEND 1974; AEND 1975b; Hanlon 1975; Anonymous 1976; Powell and Jayaremen 2002). The addition of thiotepa to village water has been denied by WHO (Tomiche 1975), but publications preceding the accusations suggest this may have happened (pp. 85, 87, Pal 1974)—and therein lies one problem. There was a lack of clear unambiguous communication from the beginning. Furthermore, PHS did have military connections and shared materials and information with the US military (Langer 1967; Treaster 1975). The US military did conduct chemical and biological tests in the Indo-Pacific; this included the release of mosquitoes off the coast of Baker Island (“Magic Sword” 1965), the release of Bacillus globigii in Oʻahu (“Big Tom” 1965), shelling sarin nerve agent in Waiākea Forest Reserve, Hawaiʻi (“Red Oak” 1967), and the dispersal of Staphylococcus aureus enterotoxin type B over Ānewetak (Enewetak) Atoll (“DTS Test 68-50” 1968). However, in all likelihood there was no military or biological warfare connections with the GCMRU (WHO 1976; Powell and Jayaremen 2002). Covert transfer of US funds to keep GCMRU going was briefly discussed with WHO (AEND 1975c; SSWDC 1975) but this was considered too risky and the US suspended funding the project. Despite denials by WHO (Tomiche 1975), the GCMRU, which was planned to extend at least until 1978, was forced to shut down prematurely in 1975 (AEND 1975d) and the project was deemed a failure (Curtis 2007).

What can be learned from this?

1. There was a clear lack of communication resulting from a reluctance of either the WHO or the US to engage the media and comment on the allegations (AEND 1974; AEND 1975e; Anonymous 1975; Tomiche 1975). This was unfortunate as it, perhaps rationally, fueled suspicions. The public perception of public perception may differ from public perception—the individual perception of public opinion is influenced by a range of factors and may not be an accurate reflection of commonly held attitudes (e.g., Mutz 1989). The idea that providing more information would undermine support conflicts with recent results that show the more informed people are of the release of genetically modified mosquitoes the more supportive they become; however, a great deal of public engagement has to be accomplished, especially for women, minorities, and people with lower education levels and lower household incomes (Ernst et al. 2015; Kolopack et al. 2015).
2. There was a perception that these experiments would not have been permitted in Western countries and that India was being used as a testing ground (Anonymous 1975; Raghavan and Jayaraman 1975). Knowledge that a technology has been effective in other countries is one factor associated with strong public support (Ernst et al. 2015). It is unfortunate that prior programs in the US, Myanmar/Burma, Tanzania, Western Africa, and France were not communicated to the Indian press (Laven 1972; WHO 1976; Curtis 2007). The perception that an international project is being conducted to avoid home country regulation should certainly be (truthfully) avoided.
3. The potential benefits of the project to the people of India was unclear (AEND 1974; Anonymous 1975). This is perhaps most tragic of all. India suffers from mosquito vectored dengue, malaria, Japanese encephalitis, Chikungunya, and lymphatic filariasis (Sharma 2015). While a balance must be struck to not over-promise results that may not be realized, the goals and potential benefits of EGE applications must also be clearly advertised.
4. There was a lack of an a priori open and frank discussion about possible misuse of the technology (Hanlon 1975). While any technology can be potentially misused by individuals or organizations, a nation's government, and especially its military, has non-humanitarian and non-conservation priorities that can potentially conflict with the goals of humanitarian and conservation projects. Regardless of the existence of an actual conflict, the perception of possible conflict does exist, which can undermine credibility (Serafino et al. 2008; Charny 2013). Fortunately today this is widely recognized and the 1978 UN ENMOD treaty (http://www.un-documents.net/enmod.htm) may prevent, depending on interpretation, military involvement in EGE technologies except perhaps for some limited applications of type 1 and 2 systems (Appendix A). The ENMOD treaty states that “Each State Party to this Convention undertakes not to engage in military … environmental modification techniques having widespread, long-lasting or severe effects … the term "environmental modification techniques" refers to any technique for changing—through the deliberate manipulation of natural processes—the dynamics, composition or structure of the Earth, including its biota …“ However, there is still a need for an open discussion about potential malicious uses, and military involvement with EGE projects should be avoided in order to encourage international trust and cooperation.

In response, Dr. B. D. Nagchaudhuri’s, physicist and scientific adviser to the Indian Ministry of Defense, recommendations were “(A) that research proposals and projects are available to the public; and (B) that pertinent records contain clear statements as to why the objective is important, what is the [Government of India's] interest, and what is the [United States Government’s] interest” (AEND 1975f) and “ministry officials must be alerted to any sensitive problems by the technical experts involved”; also, that “each collaborative project should also be approved at the ministerial or secretary level of the ministry under which the project would fall i.e. health projects - Ministry of Health, Agricultural Projects - Ministry of Agriculture, … This should also hold true whether on the Indian side or the US side” (AEND 1975g). Dr. Hanlon recommends “At the very least, there should be an open discussion of the [biological warfare] potential of such projects before they begin, so that countries can make informed choices” (p. 103, Hanlon 1975).

There is a value to compartmentalizing different aspects of a government’s actions. It seems almost self evident that funds for research are best spent by research agencies, funds for health are best spent in agencies focused on health, funds for conservation are best spent by agencies trained in and focused on conversation. Even if there were sufficient funding and resources we would not want the EPA (Environmental Protection Agency) or DOH (Department of Health) carrying out military actions; the converse is also true. We don’t want to rely on our military to carry out conservation, human health, or humanitarian actions when there are other agencies, without conflicting priorities, that can and should be doing this (Serafino et al. 2008; Charny 2013). The author has discussed EGE's and the ENMOD treaty in person with current and former members of DARPA, a research branch of the military with an interest in EGE's, and has been told that the military has to carry out high risk (in the sense of new and experimental) research because NSF (National Science Foundation) and NIH (National Institutes of Health) cannot. I disagree. Research agencies can and should also be funding higher risk, higher payoff research instead of abdicating this role to the military—and to avoid the kinds of conflicts suggested in the WHO experience in India. This is not done in the US because of historical inertia and objectively unbalanced federal budget allocations (a Department of Defense, DOD, estimated research budget of $66 billion versus$29 billion for NIH and \$6 billion for NSF in FY2015; Hourihan and Parkes 2016). Reallocating civilian research funds to civilian agencies would also free up the military to focus on military actions and capabilities.

Recent experiences with GMO’s in the Indo-Pacific
EGE’s are likely to be initially framed in terms of the GMO (Genetically Modified Organism) crop debate (Knols et al. 2007). Within Hawaiʻi, Rainbow Papaya and GMO Taro serve as contrasting examples of the interaction between social acceptance, development, and deployment of new technologies. Carica papaya was not grown in Hawaiʻi until after European contact in 1778. The papaya industry in Hawaiʻi was devastated in the 1990’s by the ringspot virus. A genetically engineered “rainbow” papaya resistant to ringspot infection was developed at Cornell University by Dr. D. Gonsalves (Ferreria et al. 2002) who was originally from Hawaiʻi. While GMO papaya is not without controversy (e.g., Harmon 2014; see Hofschneider 2016 for recent legal developments in Hawaiʻi) it is credited with rescuing the industry and is de facto widely adopted in Hawaiʻi today (e.g., Kallis 2013).

Colocasia esculenta (Taro or Kalo in Hawaiian) was brought to Hawaiʻi by the ancient Polynesians. A wide range of Kalo varieties have had a central role in traditional Hawaiian culture as a staple food crop and continues to be economically important (Whitney et al. 1939; Fleming 1994). Furthermore, Kalo is literally the brother of humans (Hāloa) in the Hawaiian creation tradition and words for family and relationships also refer to parts of the plant (Kahumoku 1980). Taro leaf blight (Phytophthora colocasiae) was introduced to Hawaiʻi in the 1900’s and has significantly impacted Kalo (Nelson et al. 2011). Work at the University of Hawaiʻi was begun to to breed resistant varieties which resulted in patents in 2002. Separately a Chinese variety of Taro was genetically modified from 2001 to 2006 with a gene from wheat to be resistant to leaf blight. This resulted in widespread public outrage and large protest rallies in 2006 that resulted in the university relinquishing its patents and issuing an indefinite moratorium on the genetic engineering of Hawaiian Kalo (Ritte and Freese 2006; CTAHR 2009).

With these cases in mind consider a potential EGE project. Culex mosquitoes were introduced to Hawaiʻi in the mid 1800’s. They vector Plasmodium relictum which is responsible for avian malaria. Many Hawaiian forest bird species, important in traditional Hawaiian culture (e.g., ʻahu ʻula, mahiole, and in Hawaiian religion), have no immunity or tolerance to P. relictum and have become extinct, with many currently threatened, as a result (Warner 1968). These two previous contrasting examples suggest that genetically modifying non-native mosquitoes to reduce the frequency of avian malaria is much more socially acceptable than the reverse: genetically modifying native Hawaiian birds to be resistant to infection by Plasmodium (although it would be worth conducting the relevant public surveys to determine this). Also, doing the research locally in Hawaiʻi is not necessarily an advantage in terms of securing broad local public support, buy-in, and acceptance (however, it is an advantage in terms of engaging the public). These are aspects that might not initially be appreciated by scientists designing EGE technologies.

Finally, concerns about ecological effects of EGE’s are associated with strong opposition to the technology (Ernst et al. 2015). There are also questions of possible, but unlikely, bioaccumulation of toxic proteins and allergenicity (Curtis 2007; Reeves et al. 2012). In addition to the four guidelines in the previous section, despite limited time and funding, we should conduct the work to have the data on hand to address these questions to the public (Curtis 2007).

Everyone has a role to play
We live in a world that is often overly self-polarizing. I am a geneticist; I entered this field because of personal interest, excitement, and challenges of the promise and potential of genetics. Unintentionally, this has become a part of my identity. When I was first exposed to protests over genetic technology it was all too easy to feel that it was also a personal attack. This is nested within the context of broader anti-scientific popular views related to climate change, evolution, renewable energy, vaccinations, etc. The natural reaction is to reflexively move in the opposite direction and argue that genetic technologies are safe, protesters don't understand the issues, etc. and be overly dismissive; a position that I may not have had initially. The difficult but essential step for growth is to try to find a middle ground and synthesize a path forward (see also NPR 2013). Right or wrong, no single perspective can do this on its own and, because of our perspectives, we are often blind to potential issues apparent to other people. It is easier to see a potential risk if you are looking for a risk instead of working toward developing a desired application of a new technology. For example, the potential of allergic reactions to genetic modifications are real and not to be dismissed (e.g., Nordlee et al. 1996), and many crops have a strong cultural significance that many people may not be aware of such as Kalo in Hawaiʻi, discussed above, or maize in Chiapas (Bellon and Brush 1994; Peralez et al. 2005; Brush and Peralez 2007). As geneticists we are in a unique position to be able to critically assess potential benefits and risks, once we perceive them, of genetic technology from a scientific perspective. It is our responsibility to embrace and communicate this rather than contributing to destructive polarization. However, it is not our job to be overly encompassing and give equal weight to all objections; we also must be willing to rationally disagree when we reason this to be the case. For example, despite claims to the contrary (Séralini et al. 2012), there is no scientific evidence that herbicide resistant maize is carcinogenic. There is a great deal of misinformation and misconceptions surrounding who would or would not benefit, and to what degree, from golden rice (Harmon 2013). Attitudes regarding GMO's are divisive, some are not based on factual evidence and can be labeled as irrational although this quickly gets complex (Stone 2010; Lynas 2013; Blancke et al. 2015; Hicks 2015); regardless, the GMO debate will continue to prove a rich subject for the analysis of the dynamics of politics, the media, framing effects, confirmation bias, social identity, information cascades, etc., for many years to come.

An area that can benefit from improvement is to incorporate this synthesis earlier into the research and development process. If individuals with different perspectives were able to directly participate in the design of a new technology, they could shape the direction in which it develops towards an outcome that might be more desirable and socially acceptable. (Recall the effect of personal experience with Bt-cotton and its adoption in India.) Often the way development of a new technology works is in incremental steps of design, troubleshooting, and research funding, to consultation and approval from regulatory agencies, to building the logistics of application and deployment. Public consultation and asking for acceptance occurs only at the end of the day, when many steps have been cast and it is more difficult and time consuming to make fundamental revisions. One possibility is to include grant support for individuals from the social sciences to be “embedded” in a biological laboratory in order to fully participate in a laboratory’s research and conduct their own research about social attitudes, context, communication, perceptions, etc., both in its own right and as a bidirectional conduit to facilitate communication, public guidance, and knowledge transfer in the development of EGE technologies (see Kolopack et al. 2015 for a highly effective example of community engagement albeit not exactly in the same form that I am proposing here). The local community can directly participate in the development of a new technology, possibly facilitating progress in a direction that is unanticipated by the researchers, funding, and regulatory agencies, but one that results in a greater positive potential being realized at the end of the day.

Conclusion
This Indo-Pacific is geographically isolated, under a burden of infectious diseases, and is in dire need of protection of its natural heritage. This creates an opportunity for positive, highly valued, effective applications of EGE's. However, it would be a mistake to ignore the history and social realities thought the region. To reiterate the eight points from the introduction that have been expanded upon through this article:

1. There is a need to enhance and engage communications in all directions.
2. EGE applications should only be pursued if there is a genuine benefit to, and buy-in from, the local population.
3. The potential benefits and risks of EGE's need to be unambiguously communicated.
4. There needs to be a clear unambiguous discussion of unintended side effects and potential misuses of the technology.
5. Humanitarian goals need to be administered and controlled by humanitarian organizations and conservation goals need to be administered and under the control of conservation organizations.
6. Proactive research needs to be conducted and the data available to address common concerns.
7. It takes broad perspectives to broadly identify potential promises and pitfalls of EGE's.
8. An early broad-based community discussion of, and involvement in, EGE development and applications should occur.

Finally, no matter how “new” a technology or situation seems, there is still much to be learned from history.

Appendix A: Types of Evolutionary Genetic Engineering
An important concept that cannot be over-emphasized is the diverse types of EGE's and their predicted effects. At the risk of oversimplification, here are four main types with an important boundary between them.

Type 1
Deleterious EGE's that are designed to be transient and removed from the population. Examples of type 1 include the “killer-rescue” system (Gould et al. 2008), genetic sterile insect technique (Horn and Wimmer 2003), and Wolbachia in cytoplasmic incompatibility population suppression applications (Laven 1967; Knipling et al. 1968). These may persist in the wild for a shorter period of time than type 2 EGE’s.
Type 2
Generic genetic modifications not designed to change in frequency over time using evolutionary principles. In general these are expected to either drift neutrally (if there is little to no effect) or be removed by natural selection. For example fluorescent proteins are often used to mark and keep track of genetic inserts; however, these proteins can have toxic effects (e.g., Liu et al. 1999; Devgan et al. 2004; Shaner et al. 2004, 2005). This tends to reduce an organism’s fitness and these modifications are not expected to persist in the wild over many generations.
Type 3
Threshold EGE’s that cannot increase in frequency when very rare but can increase in frequency and persist indefinitely once a critical frequency point is passed.

Type 3a
Thresholds that are above a frequency of one half. These include chromosomal rearrangements (Foster et al. 1972), haploinsufficient induced underdominance (Reeves et al. 2014) and possibly some forms of maternal-effect underdominance (Akbari et al. 2013).
Type 3b
Thresholds that are below a frequency of one half. This includes Wolbachia (Hoffman et al. 2011), some forms of maternal-effect underdominance (Akbari et al. 2013), and some theoretical systems (Davis et al. 2001).
Type 4
Unconditionally driving EGE’s that can invade a population from arbitrarily low frequencies. These include Medea (Chen et al. 2007), homing endonucleases (Windbichler et al. 2011), transposable elements (Caraeto et al. 1997), meiotic drive (Cha et al. 2006) and some types of CRISPR systems (Gantz and Bier 2015; DiCarlo et al. 2015; Hammond et al. 2016).

In one perspective, the most important distinction is the boundary between 3a and 3b. This predicts what will happen without human intervention (releases of modified or unmodified individuals) among multiple populations within a species due to the forces of migration and selection (Barton and Turelli 2011). Type 1—3a will tend to reduce in range and disappear (although this may take many generations for 2 and 3a) while type 3b and 4 will tend to spread and become more established (and this may occur in a small number of generations for type 4) with the concern that once widespread enough this may be irreversible. While type 3a systems might be considered "gene drive" in a broad sense the term is probably more accurate to describe type 3b and especially type 4 systems (gene drive in the strong sense). The boundary between 3a and 3b represents a balance between ease of transformation of a population and reversibility back to a transformation free state—a balance between safety and efficiency.

Some natural EGE systems in the type 4 category have been shown to be capable of moving across subspecies and species boundaries, rapidly spreading worldwide, and lowering the average fitness of a species (e.g., Eanes et al. 1988; Morita et al. 1982; Hill et al. 2016). The concern of this possibly happening due to artificial genetic engineering is not a new one (Gould et al. 2006). For example, fully functioning transposable elements have been introduced into various new species in the lab (e.g., Brennan et al. 1984; Daniels et al. 1989), and this is often done with little to no discussion of possible escape. Fortunately there are methods of building in safeguards to minimize the chance of unintended spread in the wild (Dafa’alla et al. 2006; Gokhale et al. 2014).

Additionally, it is important to keep in mind the (sometimes unexpected) effects of mutations and selection that can change the dynamics of EGE’s. For example, Y chromosome meiotic drive can be quickly suppressed by sex chromosome aneuploidy (Lyttle 1981). Arthropod species have been observed to rapidly evolve to suppress some effects of Wolbachia (Charlat et al. 2007). Type 2 EGE’s may drift at some frequency in a population by unintended contamination (e.g., Gonzales et al. 2012; Xiao et al. 2016); one concern that goes beyond this is that genetically engineered disease resistance may be adaptive, if infection by the disease has a large enough fitness cost, and the type 2 EGE may deterministically increase in frequency in the wild, essentially becoming a type 4 EGE (although to date there are not clear examples of this, e.g., Fuchs et al. 2004). Some of these unexpected effects can be detected in laboratory experiments and incorporated into the design and predictions of the EGE.

It is already a challenge to filter out misinformation and misconceptions regarding genetic modifications. The author realizes that this adds another challenge; however, the fact is there are various types of EGE's with a range of predicted effects regarding how well they can be established in the wild and how reversible they are. It is appropriate, if possible, for these dynamics to be considered and to interact with regulatory approval and public acceptance (Harmon 2014).

References
Alphey, L. (2002). Re-engineering the sterile insect technique. Insect Biochemistry and Molecular Biology, 32(10), 1243-1247.

Altrock, P. M., Traulsen, A., Reeves, R. G., & Reed, F. A. (2010). Using underdominance to bi-stably transform local populations. Journal of Theoretical Biology, 267(1), 62–75.

Altrock, P. M., Traulsen, A., & Reed, F. A. (2011). Stability properties of underdominance in finite subdivided populations. PLoS Computational Biology, 7(11).

Akbari, O. S., Matzen, K. D., Marshall, J. M., Huang, H., Ward, C. M., & Hay, B. A. (2013). A synthetic gene drive system for local, reversible modification and suppression of insect populations. Current Biology, 23(8), 671-677.

AEND: American Embassy New Delhi. (1974). NEW DE 10039: Indian press articles on WHO biogenetic mosquito control project and migratory bird project. Declassified/Released US Department of State EO Systematic Review, (July). Retrieved from https://wikileaks.org/plusd/cables/1974NEWDE10039_b.html

AEND: American Embassy New Delhi. (1975a). NEW DE 08068: WHO mosquito genetic project. Declassified/Released US Department of State EO Systematic Review, (June). Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE08068_b.html

AEND: American Embassy New Delhi. (1975b). New DE 06070: US and WHO again linked with germ warfare activities in India. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE06070_b.html

AEND: American Embassy New Delhi. (1975c). NEW DE 14272: WHO mosquito genetic control project. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE14272_b.html

AEND: American Embassy New Delhi. (1975d). NEW DE 09077: Genetic control of mosquitoes. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE12250_b.html

AEND: American Embassy New Delhi. (1975e). NEW DE 12250: Refutation of germ warfare charges. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE12250_b.html

AEND: American Embassy New Delhi. (1975f). NEW DE 09023: US-GOI Research Projects. Declassified/Released US Department of State EO Systematic Review, (July). Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE09023_b.html

AEND: American Embassy New Delhi. (1975g). NEW DE 12522: Indo-US joint commission meeting in October -- working draft of minutes. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975NEWDE12522_b.html

Anonymous. (1975). Oh, New Delhi; oh, Geneva. Nature, 256, 355–357.

Anonymous (1976) Indian attack on foreign research funding. New Scientist, 20 May, 396.

Anonymous (2016) Monsanto kicked out for giving farmers raw deal. The Times of India. Published online http://timesofindia.indiatimes.com/city/mumbai/monsanto-kicked-out-for-giving-farmers-raw-deal/articleshow/55991161.cms

Barton, N. H., & Turelli, M. (2011). Spatial waves of advance with bistable dynamics: cytoplasmic and genetic analogues of Allee effects. The American Naturalist, 178(3), E48-E75.

Basheer, S. (2016). The battle over Bt cotton. The Hindu. Published online http://www.thehindu.com/opinion/op-ed/The-battle-over-Bt-cotton/article15424211.ece

Bellon, M. R., & Brush, S. B. (1994). Keepers of maize in Chiapas, Mexico. Economic Botany, 48(2), 196-209.

Benning, T. L., LaPointe, D., Atkinson, C. T., & Vitousek, P. M. (2002). Interactions of climate change with biological invasions and land use in the Hawaiian Islands: modeling the fate of endemic birds using a geographic information system. Proceedings of the National Academy of Sciences, 99(22), 14246-14249.

Blancke, S., Van Breusegem, F., De Jaeger, G., Braeckman, J., & Van Montagu, M. (2015). Fatal attraction: the intuitive appeal of GMO opposition. Trends in Plant Science, 20(7), 414-418.

Brennan, M. D., Rowan, R. G., & Dickinson, W. J. (1984). Introduction of a functional P element into the germ-line of Drosophila hawaiiensis. Cell, 38(1), 147-151.

Brush, S. B., & Perales, H. R. (2007). A maize landscape: Ethnicity and agro-biodiversity in Chiapas Mexico. Agriculture, Ecosystems & Environment, 121(3), 211-221.

Carareto, C. M., Kim, W., Wojciechowski, M. F., O'grady, P., Prokchorova, A. V., Silva, J. C., & Kidwell, M. G. (1997). Testing transposable elements as genetic drive mechanisms using Drosophila P element constructs as a model system. Genetica, 101(1), 13-33.

Cha, S. J., Mori, A., Chadee, D. D., & Severson, D. W. (2006). Cage trials using an endogenous meiotic drive gene in the mosquito Aedes aegypti to promote population replacement. The American Journal of Tropical Medicine and Hygiene, 74(1), 62-68.

Champer, J., Buchman, A., & Akbari, O. S. (2016). Cheating evolution: engineering gene drives to manipulate the fate of wild populations. Nature Reviews Genetics, 17(3), 146-159.

Charlat, S., Hornett, E. A., Fullard, J. H., Davies, N., Roderick, G. K., Wedell, N., & Hurst, G. D. (2007). Extraordinary flux in sex ratio. Science, 317(5835), 214-214.

Charny, J. R. (2013). The U.S. Military's Expanding Role in Foreign Assistance. InterAction Policy Brief. Published Online https://www.interaction.org/files/FABB%202013_Sec16_NGOAndMilitaryRelations.pdf

Chen, C. H., Huang, H., Ward, C. M., Su, J. T., Schaeffer, L. V., Guo, M., & Hay, B. A. (2007). A synthetic maternal-effect selfish genetic element drives population replacement in Drosophila. Science, 316(5824), 597-600.

Clarke, B., Murray, J., & Johnson, M. S. (1984). The extinction of endemic species by a program of biological control. Pacific Science, 38(2), 97-104.

Clarke, T. (2002). Mosquitoes minus malaria. Nature, 419(6906), 429-430.

CTAHR: College of Tropical Agriculture and Human Resources (2009). CTAHR and Taro. Published online http://www.ctahr.hawaii.edu/oc/freepubs/pdf/CTAHR_and_taro.pdf

Curtis, C. F., 2007 Destruction in the 1970s of a research unit in India on mosquito control by sterile male release and a warning for the future. Antenna 31(4): 214-216.

Dafa'alla, T. H., Condon, G. C., Condon, K. C., Phillips, C. E., Morrison, N. I., Jin, L., ... & Alphey, L. (2006). Transposon-free insertions for insect genetic engineering. Nature Biotechnology, 24(7), 820-821.

Daniels, S. B., Chovnick, A., & Kidwell, M. G. (1989). Hybrid dysgenesis in Drosophila simulans lines transformed with autonomous P elements. Genetics, 121(2), 281-291.

Darwin, C. R. 1845. Journal of researches into the natural history and geology of the countries visited during the voyage of H.M.S. Beagle round the world, under the Command of Capt. Fitz Roy, R.N. 2d edition. London: John Murray.

Davis, S., Bax, N., & Grewe, P. (2001). Engineered underdominance allows efficient and economical introgression of traits into pest populations. Journal of Theoretical Biology, 212(1), 83-98.

Devgan, V., Rao, M. R., & Seshagiri, P. B. (2004). Impact of embryonic expression of enhanced green fluorescent protein on early mouse development. Biochemical and Biophysical Research Communications, 313(4), 1030-1036.

DiCarlo, J. E., Chavez, A., Dietz, S. L., Esvelt, K. M., & Church, G. M. (2015). Safeguarding CRISPR-Cas9 gene drives in yeast. Nature Biotechnology 33: 1250–1255

Dobson, R. (2000), Royalty-free licenses for genetically modified rice made available to developing countries, Bulletin of the World Health Organisation, 78 (10): 1281

Dobzhansky, T. (1973). Nothing in Biology Makes Sense Except in the Light of Evolution. The American Biology Teacher, 35(3), 125–129.

Eanes, W. F., Wesley, C., Hey, J., Houle, D., & Ajioka, J. W. (1988). The fitness consequences of P element insertion in Drosophila melanogaster. Genetical Research, 52(01), 17-26.

Enserink, M. (2008). Tough Lessons From Golden Rice. Science 320 (5875): 468–471

Ernst, K. C., S. Haenchen, K. Dickenson, M. S. Doyle, K. Walker, A. J. Monagen, and M. H. Hayden. 2015. Awareness and support of release of genetically modified “sterile” mosquitoes, Key West, Florida, USA. Emerging Infectious Diseases 21(2): 320-324.

Esvelt, K. M., Smidler, A. L., Catteruccia, F., & Church, G. M. (2014). Concerning RNA-guided gene drives for the alteration of wild populations. Elife, 3, e03401.

Ferreira, S. A., Pitz, K. Y., Manshardt, R., Zee, F., Fitch, M., & Gonsalves, D. (2002). Virus Coat Protein Transgenic Papaya Provides Practical Control of Papaya ringspot virus in Hawaii. Plant Disease, (February), 101–105.

Fleischer, R. C., McIntosh, C. E., & Tarr, C. L. (1998). Evolution on a volcanic conveyor belt: using phylogeographic reconstructions and K–Ar‐based ages of the Hawaiian Islands to estimate molecular evolutionary rates. Molecular Ecology, 7(4), 533-545.

Fleming, K. (1994). The Economics of Commercial Wetland Taro Production in Hawaii. Agribusiness, 7, 1–6.

Fonseca, C. R. (2009). The silent mass extinction of insect herbivores in biodiversity hotspots. Conservation Biology, 23(6), 1507-1515.

Foster, G. G., Whitten, M. J., Prout, T., & Gill, R. (1972). Chromosome rearrangements for the control of insect pests. Science, 176(4037), 875–880.

Fuchs, M., Chirco, E. M., & Gonsalves, D. (2004). Movement of coat protein genes from a commercial virus-resistant transgenic squash into a wild relative. Environmental Biosafety Research, 3(1), 5-16.

Gantz, V. M., & Bier, E. (2015). The mutagenic chain reaction: a method for converting heterozygous to homozygous mutations. Science, 348(6233), 442-444.

Ganzhorn, J. U., Lowry, P. P., Schatz, G. E., & Sommer, S. (2001). The biodiversity of Madagascar: one of the world’s hottest hotspots on its way out. Oryx, 35(4), 346-348.

Gokhale, C. S., Reeves, R. G., & Reed, F. A. (2014). Dynamics of a combined medea-underdominant population transformation system. BMC Evolutionary Biology, 14(1), 1.

Gonsalves, D., Gonsalves, C., Carr, J., Tripathi, S., Matsumoto, T., Suzuki, J., ... & Pitz, K. (2012). Assaying for pollen drift from transgenic ‘Rainbow’ to nontransgenic ‘Kapoho’ papaya under commercial and experimental field conditions in Hawaii. Tropical Plant Biology, 5(2), 153-160.

Gould, F. (2008). Broadening the application of evolutionarily based genetic pest management. Evolution, 62(2), 500-510.

Gould, F., Magori, K., & Huang, Y. (2006). Genetic Strategies for Controlling Mosquito-Borne Diseases. American Scientist, 94, 238–246.

Gould, F., Huang, Y., Legros, M., & Lloyd, A. L. (2008). A Killer–Rescue system for self-limiting gene drive of anti-pathogen constructs. Proceedings of the Royal Society of London B: Biological Sciences, 275(1653), 2823-2829.

Hammond, A., Galizi, R., Kyrou, K., Simoni, A., Siniscalchi, C., Katsanos, D., ... & Burt, A. (2016). A CRISPR-Cas9 gene drive system targeting female reproduction in the malaria mosquito vector Anopheles gambiae. Nature Biotechnology, 34(1), 78-83.

Hanlon, J. 1975. Germ-war allegations force WHO out of Indian mosquito project. New Scientist 68(970, 9 Oct) 102-103

Harmon, A. 2013. Golden Rice: Lifesaver? The New York Times, August 25, 2013, New York Edition, page SR1. Available online http://www.nytimes.com/2013/08/25/sunday-review/golden-rice-lifesaver.html

Harmon, A. 2014. On Hawaii, a Lonely Quest for Fact. The New York Times, January 5, 2014, New York Edition, page A1. Available online https://www.nytimes.com/2014/01/05/us/on-hawaii-a-lonely-quest-for-facts-about-gmos.html

Hays, W. S., & Conant, S. (2007). Biology and impacts of Pacific Island invasive species. 1. A worldwide review of effects of the small Indian mongoose, Herpestes javanicus (Carnivora: Herpestidae). Pacific Science, 61(1), 3-16.

Henneman, M. L., & Memmott, J. (2001). Infiltration of a Hawaiian community by introduced biological control agents. Science, 293(5533), 1314-1316.

Hicks, D. J. (2015). Epistemological depth in a GM crops controversy. Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences, 50, 1-12.

Hill, T., Schlötterer, C., & Betancourt, A. J. (2016). Hybrid Dysgenesis in Drosophila simulans Associated with a Rapid Invasion of the P-Element. PLoS Genetics, 12(3), e1005920.

Hoffmann, A. A., Montgomery, B. L., Popovici, J., Iturbe-Ormaetxe, I., Johnson, P. H., Muzzi, F., ... & Cook, H. (2011). Successful establishment of Wolbachia in Aedes populations to suppress dengue transmission. Nature, 476(7361), 454-457.

Hofschneider, A. (2016) Hawaii GMO Seed Companies Win Big In Appeals Court Rulings. Honolulu Civil Beat. Published online http://www.civilbeat.org/2016/11/hawaii-gmo-seed-companies-win-big-in-appeals-court-ruling/

Horn, C., & Wimmer, E. A. (2003). A transgene-based, embryo-specific lethality system for insect pest management. Nature Biotechnology, 21(1), 64-70.

Hourihan, M., & Parkes, D. (2016). Federal R&D in the FY 2016 Budget: An Overview. American Association for the Advancement of Science. Published online https://www.aaas.org/fy16budget/federal-rd-fy-2016-budget-overview

Howarth, F. G. (1983). Classical biocontrol: panacea or Pandora’s box. Proceedings of the Hawaiian Entomological Society, 24(2), 239-244.

IMF: International Monetary Fund (2016) World Economic Outlook Database. Published online http://www.imf.org/external/pubs/ft/weo/2016/01/weodata/index.aspx

Jayaraman, K. S., and C. Raghavan. 1975. The WHO and mosquitoes. Nature 257, 175-176.

Kahumoku, G. (1980). A Hawaiian Perspective on Taro Growing. Research Extension Series / Hawaiʻi Institue of Tropical Agriculture and Human Resources.

Kallis, T. 2013 Papaya: A GMO success story. Hawaii Tribune Herald, June 10. Available online http://hawaiitribune-herald.com/sections/news/local-news/papaya-gmo-success-story.html

Kathage, J., & Qaim, M. (2012). Economic impacts and impact dynamics of Bt (Bacillus thuringiensis) cotton in India. Proceedings of the National Academy of Sciences, 109(29), 11652-11656.

Kier, G., Kreft, H., Lee, T. M., Jetz, W., Ibisch, P. L., Nowicki, C., ... & Barthlott, W. (2009). A global assessment of endemism and species richness across island and mainland regions. Proceedings of the National Academy of Sciences, 106(23), 9322-9327.

Knipling, E. F., Laven, H., Craig, G. B., Pal, R., Kitzmiller, J. B., Smith, C. N., & Brown, A. W. A. 1968. Genetic control of insects of public health importance. Bulletin of the World Health Organization 38(3), 421.

Knols, B. G., Bossin, H. C., Mukabana, W. R., & Robinson, A. S. 2007. Transgenic mosquitoes and the fight against malaria: managing technology push in a turbulent GMO world. The American Journal of Tropical Medicine and Hygiene, 77(6 Suppl), 232-242.

Kolopack, P. A., Parsons, J. A., & Lavery, J. V. 2015. What makes community engagement effective?: lessons from the Eliminate Dengue Program in Queensland Australia. PLoS Neglected Tropical Disease 9(4), e0003713.

Langer, E. 1967. Chemical and Biological Warfare (I): The Research Program. Science 155(3759), 174-179.

Láruson, Á. J., & Reed, F. A. (2016). Stability of underdominant genetic polymorphisms in population networks. Journal of Theoretical Biology, 390, 156-163.

Laven, H. 1967. Eradication of Culex pipiens fatigans through cytoplasmic incompatibility. Nature 216, 383-384.

Laven, H., J. Cousserans, and G. Guille. 1972. Eradicating mosquitoes using translocations: a first field experiment. Nature 236, 456-457.

Liu, H. S., Jan, M. S., Chou, C. K., Chen, P. H., & Ke, N. J. (1999). Is green fluorescent protein toxic to the living cells? Biochemical and Biophysical Research Communications, 260(3), 712-717.

Loehle, C., & Eschenbach, W. (2012). Historical bird and terrestrial mammal extinction rates and causes. Diversity and Distributions, 18(1), 84-91.

Lynas, M. (2013). Time to call out the anti-GMO conspiracy theory. Mark Lynas speech hosted by the International Programs–College of Agriculture and Life Sciences (50th Anniversary Celebration), and the Atkinson Center for a Sustainable Future, Cornell University.

Lynas, M. (2013). The True Story About Who Destroyed a Genetically Modified Rice Crop. Future Tense: ASU | New America | Slate Published online http://www.slate.com/blogs/future_tense/2013/08/26/golden_rice_attack_in_philippines_anti_gmo_activists_lie_about_protest_and.html

Lyttle, T. W. (1981). Experimental population genetics of meiotic drive systems. III. Neutralization of sex-ratio distortion in Drosophila through sex-chromosome aneuploidy. Genetics, 98(2), 317-334.

Menon, P. (2001). Waking up to GM cotton. Frontline. 18(23). Available online http://www.frontline.in/navigation/?type=static&page=flonnet&rdurl=fl1823/18230440.htm

McGray, D. (2002). Biotech's Black Market. Mother Jones, September/October Issue. Available online http://www.motherjones.com/politics/2002/09/biotechs-black-market

Messing, R. H., & Wright, M. G. (2006). Biological control of invasive species: solution or pollution? Frontiers in Ecology and the Environment, 4(3), 132-140.

Mora, C., Frazier, A. G., Longman, R. J., Dacks, R. S., Walton, M. M., Tong, E. J., ... & Ambrosino, C. M. (2013). The projected timing of climate departure from recent variability. Nature, 502(7470), 183-187.

Morita, T., Kubota, H., Murata, K., Nozaki, M., Delarbre, C., Willison, K., ... & Gachelin, G. (1992). Evolution of the mouse t haplotype: recent and worldwide introgression to Mus musculus. Proceedings of the National Academy of Sciences, 89(15), 6851-6855.

Mutz, D. C. (1989). The influence of perceptions of media influence: Third person effects and the public expression of opinions. International Journal of Public Opinion Research, 1(1), 3-23.

Myers, N., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A., & Kent, J. (2000). Biodiversity hotspots for conservation priorities. Nature, 403(6772), 853-858.

Nelson, S., Brooks, F., & Teves, G. (2011). Taro Leaf Blight in Hawaiʻi. College of Tropical Agriculture and Human Resources, University of Hawaiʻi, PD-71

Nemana, V. (2012). In India, Genetically Modified Crops Come at a High Price. The New York Times: India Ink. Published online http://india.blogs.nytimes.com/2012/10/16/in-india-gm-crops-come-at-a-high-price/

Nordlee, J. A., Taylor, S. L., Townsend, J. A., Thomas, L. A., & Bush, R. K. (1996). Identification of a Brazil-nut allergen in transgenic soybeans. New England Journal of Medicine, 334(11), 688-692.

NPR: National Public Radio (2013). Former Anti-GMO Activist Says Science Changed His Mind. All Things Considered, January 20, 2013. Available online http://www.npr.org/2013/01/20/169847199/former-anti-gmo-activist-says-science-changed-his-mind

Paine, Jacqueline A; Shipton, Catherine A; Chaggar, Sunandha; Howells, Rhian M; Kennedy, Mike J; Vernon, Gareth; Wright, Susan Y; Hinchliffe, Edward; Adams, Jessica L (2005). Improving the nutritional value of Golden Rice through increased pro-vitamin A content. Nature Biotechnology 23 (4): 482–7.

Pal, R. 1974. WHO/ICMR programme of genetic control of mosquitoes in India. In The Use of Genetics in Insect Control, Eds. R. Pal and M. J. Whitten. North-Holland, Amsterdam.

Parry, D. (2009). Beyond Pandora’s box: quantitatively evaluating non-target effects of parasitoids in classical biological control. Biological Invasions, 11(1), 47-58.

Perales, H. R., Benz, B. F., & Brush, S. B. (2005). Maize diversity and ethnolinguistic diversity in Chiapas, Mexico. Proceedings of the National Academy of Sciences of the United States of America, 102(3), 949-954.

Pimm, S. L., Russell, G. J., Gittleman, J. L., & Brooks, T. M. (1995). The future of biodiversity. Science, 269(5222), 347.

Powell K, Jayaraman K. S. 2002. Mosquito researchers deny plotting secret biowarfare test. Nature. 419: 867.

Potrykus, I. (2001) Golden Rice and Beyond. Plant Physiology 125 (3): 1157–1161.

Raghavan, C., and K. S. Jayaraman. 1975 The WHO and mosquitoes. Nature 257: 175-176.

Reeves, R. G., Denton, J. A., Santucci, F., Bryk, J., & Reed, F. A. (2012). Scientific standards and the regulation of genetically modified insects. PLoS Neglected Tropical Diseases, 6(1), e1502.

Reeves, R. G., Bryk, J., Altrock, P. M., Denton, J. A., & Reed, F. A. (2014). First steps towards underdominant genetic transformation of insect populations. PLoS One, 9(5), e97557.

Régnier, C., Bouchet, P., Hayes, K. A., Yeung, N. W., Christensen, C. C., Chung, D. J., ... & Cowie, R. H. (2015). Extinction in a hyperdiverse endemic Hawaiian land snail family and implications for the underestimation of invertebrate extinction. Conservation Biology, 29(6), 1715-1723.

Ritte, W., & Freese, B. (2006). Haloa. Seedling, (October), 11–14.

Sehgal NK. Doubts over US in India. Nature. 1974;251:177–178.

Serafino, N. M., Dale, C., Grimmett, R. F., Margesson, R., Rollins, J., Salaam-Blyther, T., Tarnoff, C., Woolf, A. F., Wyler, L. S., Bowman, S. (2008) CRS Report for Congress: The Department of Defense Role in Foreign Assistance: Background, Major Issues, and Options for Congress. Congressional Research Service; Foreign Affairs, Defense, and Trade Division. RL34639 Available online https://fpc.state.gov/documents/organization/110406.pdf

Séralini, G. E., Clair, E., Mesnage, R., Gress, S., Defarge, N., Malatesta, M., ... & De Vendômois, J. S. (2012). RETRACTED: Long term toxicity of a Roundup herbicide and a Roundup-tolerant genetically modified maize. Food and Chemical Toxicology, 50(11), 4221-4231.

Sharma, S. (2015) You should know: 5 deadliest mosquito-borne diseases in India. Hindustan Times. Available online http://www.hindustantimes.com/columns/you-should-know-5-deadliest-mosquito-borne-diseases-in-india/story-HVE7pprbScppKyqClE5GoN.html

Shaner, N. C., Campbell, R. E., Steinbach, P. A., Giepmans, B. N., Palmer, A. E., & Tsien, R. Y. (2004). Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nature Biotechnology, 22(12), 1567-1572.

Shaner, N. C., Steinbach, P. A., & Tsien, R. Y. (2005). A guide to choosing fluorescent proteins. Nature Methods, 2(12), 905-909.

Sinkins, S. P., & Gould, F. (2006). Gene drive systems for insect disease vectors. Nature Reviews Genetics, 7(6), 427-435.

SSWDC: Secretary of State Washington DC. (1975). STATE 248935: WHO genetic mosquito control project from Dr. Ehrlich, Director, HEW/OIH. Declassified/Released US Department of State EO Systematic Review. Retrieved from https://wikileaks.org/plusd/cables/1975STATE248935_b.html

Stone, G. D. (2010). The anthropology of genetically modified crops. Annual Review of Anthropology, 39, 381-400.

Tang, G; Qin, J; Dolnikowski, GG; Russell, RM; Grusak, MA (2009). Golden Rice is an effective source of vitamin A. Am J Clin Nutr. 89 (6): 1776–83.

Tomiche, F. J. 1975 The WHO and mosquitoes. Nature 257: 175.

Treaster, J. B. 1975 U.S. Health Service Aided Army on Poison Weapon. The New York Times. September 18, p. 1 col. 7, p. 26 col. 4.

Vitousek, P. M., (1988). Chapter 20: Diversity and Biological Invasions of Oceanic Islands. pp. 181-189. In Biodiversity. Eds. Wilson, E. O., Peter, F. M., National Academies Press, Washington, D.C.

Warner, R. E. (1968). The role of introduced diseases in the extinction of the endemic Hawaiian avifauna. The Condor, 70(2), 101-120.

Webber, B. L., Raghu, S., & Edwards, O. R. (2015). Opinion: Is CRISPR-based gene drive a biocontrol silver bullet or global conservation threat? Proceedings of the National Academy of Sciences, 112(34), 10565-10567.

Whitney, L. D., Bowers, F. A. I., & Takahashi, M. (1939). Taro Varieties in Hawaii. Hawaiʻi Agricultural Experiment Station of the University of Hawaiʻi. Bulletin 84.

Wimmer, E. A. (2005). Eco-friendly insect management. Nature Biotechnology, 23(4), 432-433.

Wimmer, E. A. (2013). Insect biotechnology: controllable replacement of disease vectors. Current Biology, 23(10), R453-R456.

Windbichler, N., Menichelli, M., Papathanos, P. A., Thyme, S. B., Li, H., Ulge, U. Y., ... & Crisanti, A. (2011). A synthetic homing endonuclease-based gene drive system in the human malaria mosquito. Nature, 473(7346), 212-215.

WHO: World Health Organization (1976) WHO-supported collaborative research projects in India: the facts. WHO Chronicle, 30: 131-139.

Xiao, Q. S., Xu, W. T., Yang, J. L., & Pan, L. W. (2016). Evaluation of genetically modified rice detection methods 2011/884/EU and 2008/289/EC proposed by the European Union. Journal of Integrative Agriculture, 15(12), 2899-2910.

Ye, X; Al-Babili, S; Klöti, A; Zhang, J; Lucca, P; Beyer, P; Potrykus, I (2000). Engineering the provitamin A (beta-carotene) biosynthetic pathway into (carotenoid-free) rice endosperm. Science. 287 (5451): 303–5.

Zimmerman, H., Bloem, S., Klein, H., (2004) Biology, History, Threat, Surveillance and Control of the Cactus Moth, Cactoblastis cactorum. International Atomic Energy Agency, Austria. ISBN 92-0-108304-1. Available online: http://www-pub.iaea.org/mtcd/publications/pdf/faobsc_web.pdf

# CRISPR/Cas9 Commentary in Genetics

Late last year I was asked to write a commentary for an article that was coming out in Genetics. The article is Unckless et al. (2017) “Evolution of resistance against CRISPR/Cas9 gene drive.” The authors addressed the issue of the high mutation rates of CRISPR/Cas9 gene drive and the potential to generate alleles in the population that are resistant to the drive effects (which I have mentioned briefly before on this blog). Long story short, the evolution of resistance is essentially inevitable.The article inspired the cover image for that month (click on the image for a link and information about the artist Kent Smith).

This could be viewed as a positive thing, depending on the application, and there are already strategies being proposed to address the mutation issue. I tried to put this into perspective and suggest some alternatives in my article: Reed (2017) "CRISPR/Cas9 Gene Drive: Growing Pains for a New Technology."

# DLNR and USFWS funding!

The State of Hawaiʻi, Department of Land and Natural Resources (DLNR) and the U.S. Fish and Wildlife Service are supporting our labs' (Reed at UH Mānoa and Sutton at UH Hilo) Wolbachia project to generate incompatible mosquito mating types. This started a little while back but I didn't want to preempt the press release that DLNR was planning.

It is a huge help to get this support. We currently have a colony of Culex mosquitoes that are cleared of Wolbachia with tetracycline antibiotic. We are planning to begin microinjection of wMel Wolbachia from Drosophila fruit flies very soon. Spring break is coming up and it might be very good timing for getting a lot of work done in the lab.

Here are some links related to the press release:

http://thegardenisland.com/news/state-and-regional/birth-control-targets-hawaii-s-mosquitoes-to-protect-native-birds/article_cb3e8482-f958-11e6-8f93-4be24a81ac33.html

http://khon2.com/2017/02/22/hawaii-implements-mosquito-birth-control-to-protect-native-birds/

http://wavy.com/2017/02/23/hawaii-implements-mosquito-birth-control-to-protect-native-birds/

http://www.hawaii.edu/news/2017/02/22/birth-control-for-mosquitoes-targeted-at-saving-hawaiian-birds/

http://governor.hawaii.gov/newsroom/latest-news/dlnr-news-release-birth-control-for-mosquitoes-targeted-at-saving-unique-imperiled-hawaiian-birds/

http://www.hawaiinewsnow.com/story/34578965/state-pursues-mosquito-birth-control-to-save-native-birds

http://bigislandnow.com/2017/02/22/mosquito-birth-control-targets-saving-hawaii-birds/

http://mauinow.com/2017/02/22/mosquito-birth-control-aims-to-protect-native-birds/

https://scienceblog.com/492388/birth-control-mosquitoes-targeted-saving-hawaiian-birds/

http://www.hawaiitribune-herald.com/news/local-news/game-changer-uh-research-looking-unique-way-stop-mosquito-borne-diseases-hawaiian

http://thegardenisland.com/news/local/new-technique-could-save-endangered-birds-on-kauai/article_d8e36bf5-bd24-57a4-93f8-8db5fdc50ba5.html

http://www.kitv.com/story/34580499/birth-control-for-mosquitoes-effort-to-save-native-birds#

# $F_{ST}$ and the multinomial

The binomial distribution that many of us are more familiar with is a special case of the multinomial distribution, which can have more than two outcomes to a "test". $F_{ST}$ is a common summary of how similar populations are on a genetic level and it is based on the concept of "missing heterozygosity". I was thinking of a way to illustrate the multinomial with the probabilities of genotypes under Hardy-Weinberg assumptions, an allele frequency, $p$, and the degree of missing heterozygosity, $F$.

Here is a small dataset to work with, counts of three genotypes of a SNP with two allelic states in a sample of 30 individuals.

• AA  4
• AT  10
• TT  16

We can go back to the binomial for a moment and estimate the allele frequency. If we ignore genotype (the association between alleles), there are 18 A's and 42 T's.

• A  18
• T  42

So 30% of the alleles are A alleles. With the binomial, the probability of the data $x$ given an allele frequency $p$ is

$P(x|p) = \frac{(a+t)!}{a! t!} p^a (1-p)^t$

where $a$ is the number of A alleles and $t$ is the number of T alleles.

Plot this in R to visualize the probability of the data given the range of possible allele frequencies.

a=18
t=42
n=a+t
p <- seq(0,1,length=1000)
probx <- dbinom(a,n,p)
plot(p, probx, type='l',xlab="allele frequency", ylab="probability",
main="Binomial Probability")


We can find the maximum likelihood point by finding the partial derivative (slope) with respect to $p$, setting it equal to zero, and solving for $p$. First of all the binomial coefficient in the front is a constant so we can remove that for simplicity.

$P(x|p) \propto p^a (1-p)^t$

Let's define $n$ as the sum of $a$ and $t$.

$\frac{\partial \left( p^a (1-p)^{n-a} \right)}{\partial p}=a(1-p)^{n-a}p^{a-1}-n-a(1-p)^{n-a-1}p^a=p^{a-1}(1-p)^{n-a-1}(a-np)$

$p^{a-1}(1-p)^{n-a-1}(a-np)=0$

$p=\frac{a}{n}=\frac{18}{60}=0.3$

This makes intuitive sense. The best guess for the allele frequency is the number of times it is observed out of the total.

We can also integrate to get the area under the curve. The indefinite integral is

$\int p^a (1 - p)^{n - a} dp =\frac{ p^{a+1} \,_2F_1(a+1, a-n, 2+a, p)}{a+1}$

$\,_2F_1$ is the ordinary hypergeometric function. First this has to be evaluated at $p=1$ to put it on a scale where the area under the curve sums to one. For $a=18$ and $n=60$ this comes out to $c=1.7722\times10^{-17}$.

Evaluating

$\frac{1}{c}\frac{ p^{a+1} \,_2F_1(a+1, a-n, 2+a, p)}{a+1}|_{p=0.3}=0.4703$

at the maximum likelihood point, $p=0.3$, gives a cumulative probability of 0.4703, which is not far from 0.5, the median of the distribution. You can evaluate it online with this link to Wolfram Alpha:

https://www.wolframalpha.com/input/?i=(0.3%5E19*hypergeometric2f1(19,-42,20,0.3))%2F(19*1.7722*10%5E(-17))

Evaluating

$\frac{1}{c}\frac{ p^{a+1} \,_2F_1(a+1, a-n, 2+a, p)}{a+1}|_{p=0.5}=0.9991$

at $p=0.5$ for example gives a cumulative probability of 0.9991; so, more than 99% of the curve is below this point and we can say that $p$ is significantly less than $0.5$ given the data.

Okay, that was a warm up; now let's do this with the genotypes before adding in $F$.

Going back to the original genotype dataset of 30 individuals.

• AA  4
• AT  10
• TT  16

The multinomial probability is

$P(x|p) = \frac{(q_{aa}+q_{at}+q_{tt})!}{q_{aa}! q_{at}! q_{tt}!} \left[p^2\right]^{q_{aa}} \left[2 p (1-p)\right]^{q_{at}} \left[(1-p)^2\right]^{q_{tt}}$

where $q_{xx}$ are the counts of each genotype in the sample and $p$ is still the frequency of allele "A."

We can plot this along with the binomial curve for comparison with the following Mathematica code.

aa = 4;
at = 10;
tt = 16;
a = 2*aa + at;
t = 2*tt + at;
Plot[{Multinomial[aa, at, tt] (p^2)^aa (2 p (1 - p))^at ((1 - p)^2)^
tt, Binomial[(a + t), a] p^a (1 - p)^t}, {p, 0, 1} ,
PlotRange -> Full, PlotLegends -> {"Multinomial", "Binomial"},
AxesLabel -> {"p", "P(x)"}]


The curves are similar, they peak in the same place, but overall the multinomial curve has a lower probability than the binomial. This is because there is more information in the multinomial. Each set of the data for the binomial (allele counts) can correspond to a range of ways to divide the data into genotype counts. We can always reduce genotypes to number of alleles but we cannot do the reverse. So, a particular genotype outcome is less likely (a smaller part of all possible outcomes) than the corresponding allele count outcome.

Okay, now it gets a little more fun. We have a probability over a dimension of allele frequency of $p$ from zero to one. Let's add another dimension of $F$ but first let's think about what $F$ means. $F$ is the proportion of expected (according to Hardy-Weinberg) heterozygosity that is missing. The heterozygous genotypes don't just disappear. The alleles get reorganized into homozygous genotypes. The frequency of each allele in the heterozygotes is 1/2; so, each homozygous genotype is increased by an equal proportion, which is half of the heterozygotes that are missing. Thus, the frequency of each genotype expected given $p$ and $F$ is

$f(AA) = p^2 + F 2 p (1-p)/2 = p^2 + F p (1-p)$
$f(AT) = 2 p (1-p) - F 2 p (1-p) = 2 p (1-p) (1-F)$
$f(TT) = (1-p)^2 + F 2 p (1-p)/2 = (1-p)^2 + F p (1-p)$

This can be incorporated into the individual genotype probability components of the multinomial.

$P(x|p,F) = \frac{(q_{aa}+q_{at}+q_{tt})!}{q_{aa}! q_{at}! q_{tt}!} \left[p^2+F p (1-p)\right]^{q_{aa}} \left[2 p (1-p) (1-F)\right]^{q_{at}} \left[(1-p)^2+ F p (1-p)\right]^{q_{tt}}$

We can simply the frequency of each genotype terms by symbolizing then with corresponding $f_{XX}$ terms.

$P(x|p,F) = \frac{(q_{aa}+q_{at}+q_{tt})!}{q_{aa}! q_{at}! q_{tt}!} f_{AA}^{q_{aa}} f_{AT}^{q_{at}} f_{TT}^{q_{tt}}$

This is what the likelihood surface looks like over the range of $p$ from zero to one and $F$ from zero to one.

Here is the mathematica code used to make the plot.

a = 4;
h = 10;
t = 16;
Plot3D[Multinomial[a, h, t] (p^2 + f p (1 - p))^
a ((1 - f) 2 p (1 - p))^h ((1 - p)^2 + f p (1 - p))^t, {p, 0,
1}, {f, 0, 1}, PlotRange -> Full, AxesLabel -> {"p", "F", ""},
PlotPoints -> 100]


The curve at $F=0$ corresponds to the multinomial curve above (in the plot comparing it to the binomial). We can also see that allowing positive values for $F$ can result in higher likelihoods of the data. What do we actually expect $F$ to be? The number of heterozygotes are 10 out of a sample of 30 genotypes, so the fraction of heterozygous genotypes is 0.333. The allele frequency is 0.3, so we expect $2 p (1-p) = 2 \times 0.3 \times 0.7 = 0.42$ of the sample to be heterozygous. There is a $0.42-0.333 = 0.087$ proportion of heterozygotes that are missing. This is $0.087 / 0.42 = 0.207$ of the heterozygosity expected. So our point estimate of $F$ is 0.207, or about 21% of the heterozygosity is missing. This appears to correspond to the top of the likelihood peak. Finally, notice that the peak is narrower in the $p$ dimension than in the $F$ dimension. There is more information in the dataset about the allele frequency than there is about the genotype frequencies. There is a larger sample, 60, of alleles than there are of genotypes, 30. So, the estimate of $p$ is more refined than the estimate of $F$.

The range of zero to one makes sense for $p$. An allele frequency cannot be less than zero or greater than 1. Also, $F$ cannot be more than one (all of the expected heterozygosity is missing), but it can be less than zero (negative $F$ means there is an excess of heterozygosity). It looks like a large volume of the likely values of $p$ and $F$ given the data correspond to a negative $F$. So, $F$ is not significantly greater than zero (Hardy-Weinberg expectations) for this dataset. Let's extend the plot to encompass a range of negative values for $F$.

...and something weird happened. There is a second likelihood peak with different values of $p$ and $F$ that actually has an even higher likelihood at its peak, but the parameter values are unreasonable (given what we can see in the dataset about the allele frequency and level of heterozygosity). The minimum possible $F$ is negative one. If the entire sample were made up of heterozygotes then the allele frequency is one half, the proportion of heterozygotes expected would also be one half. This leads to a negative one half divided by one half or negative one. Extending the plot out to the full range looks even more pathological (I limited the scale and truncated the peak because it was so large).

Long story short it turns out that this is an artifact... If you closely inspect what is going on by spot checking a few points in the second peak you find that the expected number of the rarer homozygote is negative, which is impossible, and in the likelihood calculation this is raised to the fourth power, which makes it positive, etc. This is a good example of how unanticipated, and initially cryptic, errors can creep into a computational script. This is fixed by setting up an if statement to only evaluate the likelihood if the expected number of homozygotes are positive; otherwise the likelihood is zero, which is true. Here is the updated script and a plot.

a = 4;
h = 10;
t = 16;
Plot3D[Multinomial[a, h, t] If[
p^2 + f p (1 - p) >
0, (p^2 + f p (1 - p))^a ((1 - f) 2 p (1 - p))^
h ((1 - p)^2 + f p (1 - p))^t, 0], {p, 0, 1}, {f, -1, 1},
PlotRange -> Full, AxesLabel -> {"p", "F", ""}, PlotPoints -> 100]


Because $F$ and $p$ appear uncorrelated we can feel good about taking a slice of the likelihood surface through the maximum-likelihood value of $p$ as a reasonable approximation of the (contour) likelihood surface of $F$ in order to evaluate the range of possible parameter values.

This is not yet $F_{ST}$. We have been evaluating the missing heterozygosity of a single population sample. Heterozygosity can be reduced within a population by inbreeding effects where individuals do not mate randomly but tend to mate more often with closer relatives (selfing in plants is an extreme example). What was calculated above is $F_{IS}$.

Let's say we had another sample from a different population. So now we have our original population one

• AA 4
• AT 10
• TT 16

and population two.

• AA 10
• AT 14
• TT 6

Plotting the likelihood surfaces together.

a1 = 4;
h1 = 10;
t1 = 16;
a2 = 10;
h2 = 14;
t2 = 6;
Plot3D[{If[p^2 + f p (1 - p) > 0 && (1 - p)^2 + f p (1 - p) > 0,
Multinomial[a1, h1, t1] (p^2 + f p (1 - p))^
a1 ((1 - f) 2 p (1 - p))^h1 ((1 - p)^2 + f p (1 - p))^t1, 0],
If[p^2 + f p (1 - p) > 0 && (1 - p)^2 + f p (1 - p) > 0,
Multinomial[a2, h2, t2] (p^2 + f p (1 - p))^
a2 ((1 - f) 2 p (1 - p))^h2 ((1 - p)^2 + f p (1 - p))^t2, 0],
0.0005}, {p, 0, 1}, {f, -1, 1}, PlotRange -> Full,
AxesLabel -> {"p", "F", ""}, PlotPoints -> 100]


The second sample has a higher allele frequency and a value of $F_{IS}$ closer to zero (however, the two ranges of likely $F$ values overlap and might not be considered significantly different). The second allele frequency is $p_2 = (10+14/2)/30 = 0.567$. With the two allele frequencies we can now calculate a point estimate of $F_{ST}$ between the population samples. The expected heterozygosity in sample 2 is $H_{S,2} = 2 \times 0.567 \times (1-0.567) = 0.491$. We already showed above that $p_1 = 0.3$ and $H_{S,1} = 0.42$. The average allele frequency over both populations is $\bar{p} = 0.434$ and the average subpopulation expected heterozygosity is $\bar{H}_S = 0.456$. In a combined total sample, if there were population structure, the expected heterozygosity is $H_T = 2 \bar{p} (1-\bar{p}) = 0.491$. Therefore,

$F_{ST} = \frac{H_T - \bar{H_S}}{H_T} = \frac{0.491-0.456}{0.491} = 0.0713$

or about 7% of the expected heterozygosity is missing because of differences in allele frequencies between the populations. Why does this work? Any time there is a difference in allele frequency the average subpopulation expected heterozygosity will be lower than the combined expected heterozygosity. Why? Because the curve of heterozygosity over allele frequency space is a concave function; the average of two points will always be lower than the value of the function evaluated at that point.

$F_{ST}$ is the gap between $H_T$ and $H_S$, which is 7% of $H_T$.

Before bringing this back together I need to introduce one more $F$-statistic. This one is $F_{IT}$. It is the total missing heterozygosity due to both population structure and inbreeding effects. If we combine the two samples we get

• AA 14
• AT 24
• TT 22

The allele frequency $p_T = 0.433$ (the same as the average between the two samples) and the fraction of heterozygotes in the combined sample is $24/60=$ 40%. We expect a heterozygosity of $2 p_T (1-p_T) = 0.491$. So, $F_{IT} = (0.491-0.4)/0.491 = 0.185$.

$F_{ST}$ between populations is related to $F_{IT}$ and $F_{IS}$. If we know the latter two we can estimate the former.

$F_{ST} = \frac{F_{IS}-F_{IT}}{F_{IS}-1}$

Using an $F_{IT}=0.185$ and let's say an $F_{IS}=0.125$ we get

$F_{ST} = \frac{0.125-0.185}{0.125-1} = \frac{0.06}{0.875} = 0.0686$.

We can estimate $F_{IT}$ and $F_{IS}$ using the multinomial, which means we can also estimate $F_{ST}$. This is starting to get more complicated with lots of parameters to estimate from the data, two allele frequencies and two $F$ statistics (if we have a single $F_{IS}$ over both populations, which seems biologically reasonable) or four parameters in total. A good way to explore higher dimensional parameter space is to use a Markov Chain in a Monte Carlo fashion (MCMC) with Metropolis-Hastings algorithm. I save the details of this method for a later post. For now let's write a "hill climbing" Monte Carlo script (Monte Carlo refers to guessing parameter values and hill climbing means that it moves to values that produce higher probabilities of the data). Here it is in R.

# data
pop1=c(4,10,16)
pop2=c(10,14,6)
popt = pop1 + pop2

# intialize variables
probbest = 0
p1best=0.5
p2best=0.5
fisbest=0
fitbest=0

# loop through some steps
for (i in 1:100000) {

tune = 1/i # narrows in as the run goes

# pick new parameter values and check to make
# sure they are not out of bounds
p1 = p1best + runif(1, min = -tune, max = tune)
if (p1<0) {p1=0} else if (p1>1) {p1=1}
p2 = p2best + runif(1, min = -tune, max = tune)
if (p2<0) {p2=0} else if (p2>1) {p2=1}
fis = fisbest + runif(1, min = -tune, max = tune)
if (fis< -1) {fis= -1} else if (fis>1) {fis=1}
fit = fitbest + runif(1, min = -tune, max = tune)
if (fit< -1) {fit= -1} else if (fit>1) {fit=1}
pt = (p1+p2)/2

# genotype probabilities
gp1=c(p1^2+fis*p1*(1-p1),2*p1*(1-p1)*(1-fis),(1-p1)^2+fis*p1*(1-p1))
gp2=c(p2^2+fis*p2*(1-p2),2*p2*(1-p2)*(1-fis),(1-p2)^2+fis*p2*(1-p2))
gpt=c(pt^2+fit*pt*(1-pt),2*pt*(1-pt)*(1-fit),(1-pt)^2+fit*pt*(1-pt))

# sample probabilities
if(p1^2+fis*p1*(1-p1)>0 & (1-p1)^2+fis*p1*(1-p1)>0) {
prob1<-dmultinom(pop1, , gp1)
} else {
prob1 = 0
}
if(p2^2+fis*p2*(1-p2)>0 & (1-p2)^2+fis*p2*(1-p2)>0) {
prob2<-dmultinom(pop2, , gp2)
} else {
prob2 = 0
}
if(pt^2+fit*pt*(1-pt)>0 & (1-pt)^2+fit*pt*(1-pt)>0) {
probt<-dmultinom(popt, , gpt)
} else {
probt = 0
}
# combined probabilities
prod = prob1 * prob2 * probt

# keep the best ones
if(prod>probbest){
p1best=p1
p2best=p2
fisbest=fis
fitbest=fit
probbest=prod
} else {
p1 = p1best
p2=p2best
fis = fisbest
fit = fitbest
}
pt = (p1+p2)/2

# calculate fst
fst = (fis-fit)/(fis-1)

# report results
print(c(p1,p2,fis,fit,fst))
}


Running this for 100,000 steps I got estimates of $p_1 = 0.299$, $p_2 = 0.567$, $F_{IS} = 0.125$, $F_{IT} = 0.185$, and $F_{ST} = 0.0687$; not bad.

Okay, that's going to have to be it for this post. This could be extended to cases where the sample sizes are unequal, multiple loci ($F$ is shared across loci but $p$ is not), more than two alleles, more than two populations and hierarchical population groupings, and an MCMC script to estimate the probability distribution of $F_{ST}$, but I'll have to save that for later.

# Mosquito Busters!

This weekend I teamed up with Matt Medeiros to collect mosquitoes. We were looking for adult Culex and Aedes. Culex larvae are easy to find in large numbers but the adults are difficult to locate.  We did find a lot of Aedes albopictus and brought them back to the lab for dissection and DNA isolation.

Matt had a backpack vacuum to catch the adults. In the image above we are looking for mosquitoes resting on the walls in a tunnel system under Honolulu (the image is fuzzy because of the low light levels and being hand held). It is impossible to resist Ghostbusters coming to mind.

In other, not unrelated, news we seem to have been too heavy handed with the tetracycline to clear Wolbachia from the mosquitoes. Most of the mosquitoes in the tetracycline cage died (presumably from the toxic effects of the antibiotic) and we did not recover any eggs.  We are trying again with a shorter term dose.

# Wolbachia PCR and Sequencing

We need a source of Wolbachia for injection into the mosquitoes. Using primers from Simōes et al. (2011) Áki and Maria were able to get these PCR products.

The gel is shown sideways with shorter fragments to the left. Our white[1] stock of Drosophila melanogaster is positive for Wolbachia infection. The Oregon-R stock appears to be uninfected. The Culex mosquitoes collected locally here on Oʻahu also appear to be infected and give a double band. We submitted them for sequencing and the Culex sequence came back very messy---consistent with possible super infection of multiple strains. However, the Drosophila w[1] sequence was clear enough to get a basepair sequence.

Here is the full sequence we recovered.

>Dmelwol_w[1]
TAATACGGAGAGRGCTAGCGTTATTCGGAATTATTGGGCGTAAAGGGCGCGTAGGCGRATTAGTAAGTTAAAAGTGAAATCCCAAGGCTCAACCTTGGAATTGCTTTTAAAACTGCTAATCTAGAGATTGAAAGAGGATAGAGGAATTCCTAGTGTAGAGGTGAAATTCGTAAATATTAGGAGGAACACCAGTGGCGAAGGCGTCTATCTGGTTCAAATCTGACGCTGAGGCGCGAAGSKTGGGGAGCAAACAGGATTAGATACCCTGGTAGTCCACGCTGTAAACGATGAATGTTAAATATGGGAAGTTTTACTTTCTGTATTACAGCTAACGCGTTAAACATTCCGCCTGGGGACTACGGTCGCAAGATTAAAACTCAAAGGAATTGACGGGGACCCGCACAAGCGGTGGAGCATGTGGTTTAATTCGATGCAACSCGAAAAACCTTACCACTCCTTGACATGGAAATTATACCTATTCGAAGGGATAGGGTCGGTTCGTCCGGGTTTCACACAGGTGTTGCATGGCTGTCGTCAGCTCGTGTCGTGAGATGTTGGGTTAAGTCCCGCAACGAGCGCAACCCTCATCCTTAGTTACCATCAGTTAATGCTGGGGACTTTAAGGAAACTGCCAGTGATAAACTGGAGGAAGGTGGGGATGATGTCAAGTCATCATGGCTATGAGTGCTACACACGTGCTACAATGGTGGCTACAATGGGCTGCAAAGTCGCGAGGCTAAGCTAATCCCTTAAAGGCCATCTCAGTCCGGATTGTACTCTGCAACTCGAGTGCATGAGTTGATCGCTAGTAATCGTGGATCAGCACGCCACGGTGATAGCTTTCTCGGGTCTG

This appears to be a Group A Wolbachia and is consistent with wMel.

Simōes, P. M., Mialdea, G., Reiss, D., Sagot, M. F., & Charlat, S. (2011). Wolbachia detection: an assessment of standard PCR protocols. Molecular Ecology Resources, 11(3), 567–572.