Keller Lab

The Keller Lab 2017


Matt Keller, Richard Border (grad student), Luke Evans (postdoc), Rasool Tahmasbi (postdoc), Emma Johnson (grad student).

The Keller Lab 2013


Nikhitha Reddy (programmer), Tess Adams (RA), Doug Bjelland (postdoc), Courtney Hibbs (RA), Matt Keller, Amanda Wills (grad student), Emma Johnson (grad student), Teresa de Candia (grad student), Dorian Mitchem (grad student)

Pictures from the 2013 Keller Lab party!

Welcome! Our lab is interested in finding motivated, quantitatively-minded graduate students or postdocs to work with. As much of our work has become programming intensive, we are also interested in finding students or researchers with computer programming and statistical experience. That said, most graduate students who work in our lab started with no programming experience and basic statistical knowledge; they just had a burning desire to answer interesting questions and learned the programming and statistics along the way. So if those qualities apply to you and you’re shopping for a lab to work in, read on...

What type of work might graduate students / postdocs do here?
Graduate students interested mostly in working with me should apply to the Behavioral Genetics (BG) area in Psychology and Neuroscience. Should you be accepted here, you will probably spend a lot of your research time at the Institute for Behavioral Genetics - a world-class research institute affiliated with both Psychology and IPHY (among others). The advantage of this is that you can work with different people on different projects - you aren’t pigeonholed into one particular lab.

Obviously we would be most interested in working with students who share some of our interests above. For both of us, it would be great to find students who are eager to learn statistical packages (Mx, PLINKseq, etc.) & languages (R, C++, etc.) used in analysis of genetic data. With that said, we should take a step back. We are really just interested in working with talented students on a broad range of projects, from topics in evolutionary psychology and psychiatry to those in behavioral and statistical genetics. The most important issue is: Does our research have a clear question and will our approach further our understanding of that question?

Why should you apply to work here?

  • CU Psychology is consistently ranked within the top 25 of psychology programs. As a consequence, it is highly selective (accepting under 10% of applicants), attracting some of the best and brightest students in psychology.
  • Institute for Behavioral Genetics is a world-class institute. By working here, you will have access to large, genetically informative datasets (for example, longitudinal datasets of adopted children or twins). Access to such datasets can make a huge difference in your publication record as a graduate student!
  • IBG students also have the option of developing skills in sophisticated molecular genetics laboratories and of learning techniques for analyzing such data.
  • Most graduate students in psychology at CU receive full financial support.
  • Boulder is a fantastic town, not too big and not too small, and often shows up on one of those “best places to live” lists. It has great public transport and a terrific bike-trail system (no problem getting around for those without vehicles). It is also one of the sunniest places in the U.S. (see one of my publications about the effects of weather on mood and cognition).
  • Lots to do: skiing, snowmobiling, and snowshoeing in winter; hiking, biking, camping, and rafting in summer/fall.
  • The Boulder campus is within walking distance of beautiful Rocky Mountain hiking and biking trails
  • And last but not least... you can help us develop the nascent field of evolutionary behavioral genetics and learn about the genetic underpinnings of psychiatric disorders..

Current and Previous Members of the Keller Lab:

Graduate Students & Postdocs
The Keller Lab holds weekly lab meetings where we discuss problems, new approaches, progress, etc. on the projects lab members are working on. Typically, each student works on one to two projects at a time, and we aim to make discoveries and see these published in prominent journals. I work closely with my graduate students to try to teach them the analytical, statistical, methodological, writing, and communication skills it takes to succeed in academia. But the mentorship has worked in both directions: my students have taught me a great deal about the field and, invariably, leave my lab with a deeper understanding of topics in the field than I have myself. I believe I have been blessed to have been a mentor and colleague of such a strong group. Below, I briefly discuss the achievements and projects of each of the past and current graduate students of the lab.

Laramie Duncan, a PhD student in the Clinical Psychology area, began working with me in 2008, after her two previous advisors had left to take positions elsewhere. Laramie first helped me work on the development of new extended twin family models (Keller, Medland, Duncan, Hatemi, Neale, Maes, & Eaves, 2009; Keller, Medland, & Duncan, 2010). For her dissertation (May 2010), Laramie critically reviewed all candidate gene-by-environment interaction research in psychiatry conducted to date. The paper that ultimately came of this work (Duncan & Keller, 2011) was published in the prestigious American Journal of Psychiatry. In it, we argued that the false positive rate of published findings in the field may be much higher than the nominal type I error rate of .05 because sample sizes have typically been small, the appropriateness of multiple testing corrections has been difficult to verify, and the unpublished ‘file drawer’ of negative findings may be large. This paper has garnered a good deal of attention, and at least two journals (Behavior Genetics and the Journal of Abnormal Child Psychology) have used our findings among others to justify new policies outlining stricter criteria that must be met before manuscripts reporting candidate gene main effects or interactions will be considered for review. Laramie is currently a postdoctoral fellow at the Broad Institute of Harvard/MIT, where she is continuing research on psychiatric genetics. We plan to co-author a chapter together in the upcoming book, Psychological Science Under Scrutiny: Recent Challenges and Proposed Solutions.

Dan Howrigan, a PhD student in the Behavioral Genetics area, began working with me in 2008. Along with Matt Simonson, he was the first student that applied and was accepted to work with me (and co-mentor Matt McQueen). Dan’s work centered on my primary research interest at the time: using whole-genome single nucleotide polymorphism (SNP) data to deduce the degree of genomic inbreeding (autozygosity) in individuals, and using that information to predict psychiatric traits. His paper comparing methods of detecting autozygosity (Howrigan, Simonson, & Keller, 2011) was an impressive piece of work, and has been used by myself and others in the field to better quantify this variable. We used that work to detect autozygosity in the largest schizophrenia case-control genetic dataset collected to that point (n~22,000) and found that it was a risk factor for schizophrenia (Keller, Simonson, Ripke, Neale, Gejman, Howrigan, Lee, Lencz, Levinson, Sullivan, & the PGC, 2012). He is also first author on a manuscript to be submitted soon testing the same hypothesis with respect to IQ; in so doing, we have put together the largest genetic dataset on IQ yet collected (n~9,000). In addition, Dan is first author on a paper we published in BMC Genetics (Howrigan, Simonson, Kamens, Stephens, Wills, Ehringer, Keller, & McQueen, 2011) and co-author on a chapter we published on evolutionary behavioral genetics (Keller, Howrigan, & Simonson, 2010). Dan successfully defended his dissertation in July of 2012. Along with Laramie, Dan is currently a postdoctoral fellow at the Broad Institute of Harvard/MIT, where he is lead analyst on a large (n~40,000) Psychiatric Genomics Consortium study investigating the effects of copy number variants on schizophrenia risk.

Matt Simonson, a PhD student in the Behavioral Genetics area, also began working with me in 2008. Matt became an expert on polygenic approaches to dissecting complex traits (Simonson, Wills, Keller, & McQueen, 2011) as well as on statistical methods for detecting genetic pathways important to disease (Simonson, McQueen, & Keller, under review). This latter publication is impressive, having detected several novel pathways affecting body mass index in a sample of 132,000 individuals. Matt also helped develop techniques to understand whether particular areas of the genome, when autozygous, are particularly predictive of phenotype, and this work has been used in several publications (Keller, Simonson, Ripke, Neale, Gejman, Howrigan, Lee, Lencz, Levinson, Sullivan, & the PGC, 2012; Howrigan, Simonson, & Keller, 2011). Matt defended his dissertation in May of 2013, and is currently a postdoctoral fellow at the Mayo Clinic in Rochester, MN, where he is working on the genetics of bipolar disorder.

Teresa de Candia, a PhD student in the Behavioral Genetics area, began working with me in 2010. Teresa has been interested in understanding the genetic architecture of schizophrenia. She has used mixed linear effects models to estimate the additive genetic variation in schizophrenia tagged by common SNPs, which was published in Nature Genetics (Lee, de Candia, Ripke, Yang, PGC, ISC, MGS, Sullivan, Goddard, Keller, Visscher, & Wray, 2012) and first author on a recently accepted paper to American Journal of Human Genetics that, for the first time, estimated the degree of genetic overlap for a trait (schizophrenia in this case) between populations of European and African descent (de Candia, Lee, Yang, Browning, Gejman, Levinson, Mowry, Hewitt, Goddard, O’Donovan, Purcell, Posthuma, ISC, MGS, Visscher, Wray, Keller, in press). This publication is particularly important because there has been concern that GWAS results might be European-centric, given that most samples are of European origin. Teresa’s paper not only shows that there is substantial overlap of schizophrenia risk alleles across European and African descent populations, it also demonstrates a method that can be used to get at this question between any ethnicities for any trait. Teresa is currently working on a project that explores which factors can cause a reduced genetic correlation between ethnicities. She plans to defend her dissertation in the spring of 2014.

Amanda Wills, a PhD student in the Behavioral Genetics area, is co-mentored with Greg Carey and I. While she has long attended my lab meetings, we began working on our first project earlier this year. Amanda has been interested in understanding the degree to which the same genetic effects that predict smoking initiation are also responsible for smoking addiction. She has previously used twin designs to get at this question, but she and I have recently used whole-genome SNP data to attempt to gain traction on this issue. She presented her preliminary results this year at the Behavioral Genetics Association meeting in Marseille (Wills & Keller, 2013). She has also helped on two other publications I have been involved with (Simonson, Wills, Keller, & McQueen, 2011; Howrigan, Simonson, Kamens, Stephens, Wills, Ehringer, Keller, & McQueen, 2011). She plans to defend her dissertation in the spring of 2014.

Dorian Mitchem, a PhD student in the Behavioral Genetics area, was a lead RA in my undergraduate lab, and impressed me enough that I offered him a graduate position in my lab. He began working with me in the fall of 2011. Dorian has been interested in understanding the genetic correlation between traits related to Darwinian fitness. He recently had a paper accepted in Behavior Genetics (Mitchem, Purkey, Grebe, Carey, Garver-Apgar, Bates, Arden, Hewitt, Medland, Martin, Zeitsch, & Keller, accepted) quantifying the sex-specific effects of facial attractiveness and sexual dimorphism. Another manuscript will be resubmitted to Psychological Science (Lee, Mitchem, Wright, Martin, Keller, & Zietsch, under review). He is currently working on another manuscript that shows that, counter to evolutionary psychology theories and our own preconceptions, there is no correlation between IQ and facial attractiveness. Finally, he is beginning to work on measured genetic datasets, trying to understand whether schizophrenia risk has genetic overlap with creativity.

Emma Johnson is a first year PhD student in the Behavioral Genetics area. She will be leading a project investigating runs of homozygosity in the Psychiatric Genomics Consortium 2 dataset

Doug Bjelland is a first year postdoctoral fellow at the Institute for Behavioral Genetics. He received his PhD in Animal Breedung at Genetics from the University of Wisconsin in 2013. He is currently helping in the development of a new method to detect identical by descent (IBD) shared haplotypes between people using SNP data. We plan on doing future work on identifying haplotypes that carry highly penetrant, recessive mutations.


Undergraduate RAs:

Alicia Purkey was a lead RA for me in the 2011-2012 year. She completed an honor’s thesis with me in the fall of 2011. She won the Imogene Jacobs Scholarship in Psychology award in May of 2011. Her thesis, The heritability of facial attractiveness, used a twin design to understand the degree to which rated facial attractiveness was heritable (it is). This was an impressive endeavor for an undergraduate, as it entailed learning the OpenMx and R statistical programs, and conducting the appropriate analyses on the data. She defended her thesis November of 2011 and received our department’s highest honor, Summa Cum Laude. This work, which was taken over and extended by my graduate student, Dorian Mitchem, has since been accepted to the Behavior Genetics journal (Mitchem, Purkey, Grebe, Carey, Garver-Apgar, Bates, Arden, Hewitt, Medland, Martin, Zeitsch, & Keller. Estimating the sex-specific effects of genes on facial attractiveness and sexual dimorphism, accepted, Behavior Genetics.). Alicia is currently a graduate student in Biomedical Sciences at the University of Colorado, Denver.

Tess Adams was a lead RA for me in the fall of 2012. She completed her honor’s thesis with me in the fall of 2012. The thesis, Judging a book by its cover: Are first impressions accurate?, found that simply by looking at a facial photograph, judges can guess at a target’s true IQ and extraversion at levels far beyond chance (p < 1 in a million). This result held up even after controlling for numerous potential mediators, such as gender, SES, attractiveness, grooming, photo quality, etc. The work employed rather advanced statistical models (mixed linear effects and structural equation models) to accurately model the dependencies induced by the family structure in the data. Her thesis was also awarded our department’s highest honor, Summa Cum Laude. She presented this work as an oral presentation at the Behavior Genetics Association meeting in Marseille France in 2013, and is currently writing up the results for submission to a psychology journal. She is an excellent writer, and I have no doubt her paper will be accepted at a high impact journal. She is planning on applying to graduate school next year.

Courtney Hibbs has been a lead RA for me since the fall of 2012 (she and Tess were co-lead RAs in fall 2012). She has done an amazing job keeping the lab organized and running. She plans to defend her honors thesis in the spring of 2014. Her thesis will further investigate how people can glean accurate information from faces, whether it is something in the facial morphology itself or stems from some other less interesting source (such as a confounds that we have not yet uncovered).



[Matthew C Keller's Home Page] [Biosketch] [Vita] [Publications] [Keller Lab] [LabParty2013] [Program Code] [R] [Courses] [Links]