Afterword: CEHG Genetics and Society Symposium 2015

CEHG_Logo_Mono_Black

Founded in 2012, CEHG is a research program that fosters interdisciplinary research. Home to more than 25 faculty and more than 200 grads and postdocs, CEHG bridges the divides between various member labs across Stanford campus.

The 2015 CEHG Genetics and Society Symposium (GSS15), which took place on April 13th and 14th in Stanford’s Paul Brest Hall, was a smashing success. It featured 25 speakers from Stanford campus and the San Francisco Bay academic and scientific industry communities. Approximately 175 Stanford affiliates and non-affiliates came together to celebrate the Center’s spirit of interdisciplinary collaboration and meet with experts in the fields of computational, evolutionary and human genomics This is a significant increase from last year’s 150 attendees!

The Mission:

The Genetics and Society Symposium is integral to CEHG’s mission: it provides postdocs and graduate fellows with the opportunity to share their developing research with faculty advisors and their colleagues, encourages conversation between faculty working in diverse scientific disciplines across campus, and introduces CEHG members to speakers from around the Bay Area and beyond (and vice versa).

The Venue:

As you can see from our photos of the space and catering service, Paul Brest Hall was the perfect home for this year’s two-day symposium. The hall was spacious, the food delicious, the staff hands on, and the outdoor picnic area well suited for our lunch and coffee breaks. We enjoyed the venue so much, in fact, that CEHG staff are currently in the process of booking the space for next year!

The Speakers:

GSS15 featured four brilliant keynote speakers, each distinguished in his/her field of research.

Gene Myers and CEHG Exec Committee members Marc Feldman, Chiara Sabatti, and Carlos Bustamante

Gene Myers and CEHG Exec Committee members Marc Feldman, Chiara Sabatti, and Carlos Bustamante

Founding director of a new Systems Biology Center at the Max-Planck Institute of Molecular Cell Biology and Genetics, Dr. Eugene (Gene) Myers presented his open-sourced research on the resurrection of de novo DNA sequencing. Best known for the development of BLAST, the most widely used tool in bioinformatics and the assembler he developed at Celera that delivered the fly, human, and mouse genomes in a three-year period, Dr. Myers participated in GSS15, courtesy of DNAnexus. Follow his blog: https://github.com/thegenemyers.

Co-founding director Carlos Bustamante and Ed Green catch up during a break at GSS15.

Co-founding director Carlos Bustamante and Ed Green catch up during a break at GSS15.

Assistant Professor in Biomolecular Engineering at the University of California, Santa Cruz, Richard (Ed) Green presented his research on a novel approach for highly contiguous genome assemblies, which draws on his work as an NSF Fellow at the Max Planck Institute in Leipzig, Germany and head of an analysis consortium responsible for publishing the draft genome sequence of Neanderthal. Click here for his 2014 CARTA talk, “The Genetics of Humanness: The Neanderthal and Denisovan Genomes.

Dr. Michelle Mello, Stanford Law School and School of Medicine

Dr. Michelle Mello, Stanford Law School and School of Medicine

Michelle Mello, Professor of Law at Stanford Law School and Professor of Health Research and Policy in Stanford’s School of Medicine, presented findings from her extensive research on the ethics of data sharing. As the author of more than 140 articles and book chapters on the medical malpractice system, medical errors and patient safety, public health law, research ethics, the obesity epidemic, and pharmaceuticals, Dr. Mello provided a valuable perspective from the intersections of law, ethics, and health policy. Click here to read Dr. Mello’s SLS profile.

Dr. Ami Bhatt, Stanford Medicine

Dr. Ami Bhatt, Stanford Medicine

Ami Bhatt shared her passion for improving outcomes for patients with hematological malignancies in her talk, “Bugs, drugs, and cancer.” Best known for her recent work demonstrating the discovery of a novel bacterium using sequence-based analysis of a diseased human tissue, her research has been presented nationally and internationally and published in 2013 in the New England Journal of Medicine. Click here for links to Dr. Bhatt’s CAP profile and lab homepage.

 

We had a large group of CEHG faculty members at this year’s event, showcasing the cutting edge research being done in CEHG labs across Stanford campus and indicating considerable faculty commitment to ensuring the Center’s continuing success.

Our symposium would not be complete without our invited CEHG Fellows. These speakers were nominated by organizing committee members to present on topics relating to their CEHG-funded research projects. These young scholars embody CEHG’s continuing commitment to provide funding support to researchers as they transition from graduate studies to postdoctoral scholarships.

The Workshop:

There was standing room only when facilitators Chiara Sabatti (Associate Professor of Health Research and Policy at Stanford), Ken Lange (Chair of the Human Genetics Department at UCLA), and Suyash Shringarpure (postdoctoral scholar in Stanford’s Bustamante Lab) presented their approaches to contemporary problems in statistical genetics!

Social Media:

Did you know? CEHG is on social media!

GSS15 social media moderators, Bridget Algee-Hewitt, Jeremy Hsu, Katie Kanagawa, and Rajiv McCoy were posting live throughout both days of the event. And our efforts to reach the larger community paid off, with a total reach of 815 on Facebook and more than 7,000 impressions on Twitter!

To catch up on our GSS15 coverage, check out our Facebook page at https://www.facebook.com/StanfordCEHG?ref=hl and our Twitter feed @StanfordCEHG. Follow both to make sure you are the first to know when we post CEHG-related news and announcements.

Want to know when speaker videos from the symposium will be available on CEHG’s forthcoming youtube channel? Follow us on Facebook and Twitter!

Special Thanks:

From left to right: Bridget Algee-Hewitt, Cody Sam, Yang Li, Anand Bhaskar, and Katie Kanagawa

From left to right: Bridget Algee-Hewitt, Cody Sam, Yang Li, Anand Bhaskar, and Katie Kanagawa

The GSS15 organizing committee—including Bridget Algee-Hewitt, Anand Bhaskar, Katie Kanagawa, Yang Li, and Cody Sam—would like to take this opportunity to thank CEHG Directors Carlos Bustamante and Marc Feldman, Executive Committee members Hank Greely, Dmitri Petrov, Noah Rosenberg, and Chiara Sabatti, event volunteers Alex Adams, Maude David, and Chris Gignoux, event photographer Deneb Semprum, and everyone who attended this year’s symposium.

We hope you enjoyed attending as much as we enjoyed working behind-the-scenes. We hope to see you all again at GSS16! If you are interested in volunteering for future CEHG events, please contact us at stanfordcehg@stanford.edu.

Upcoming CEHG events:

Don’t miss our popular weekly Evolgenome seminar series, which will continue through Spring term, usually on Wednesdays at noon (location varies). Lunch is always provided. Details will follow, but here is a quick overview so you can mark your calendars!

April 29: Fernando Racimo (Nielsen/Slatkin Lab)
May 6: Pleuni Pennings (UCSF)
May 20: Kelly Harkin
June 3: Sandeep Ventakaram (Petrov Lab)
June 10: Emilia Huerta-Sanchez

A fast and accurate coalescent approximation

Blog author Suyash Shringarpure is a postdoc in Carlos Bustamante's lab.

Blog author Suyash Shringarpure is a postdoc in Carlos Bustamante’s lab. Suyash is interested in statistical and computational problems involved in the analysis of biological data.

The coalescent model is a powerful tool in the population geneticist’s toolbox. It traces the history of a sample back to its most recent common ancestor (MRCA) by looking at coalescence events between pairs of lineages. Starting from assumptions of random mating, selective neutrality, and constant population size, the coalescent uses a simple stochastic process that allows us to study properties of genealogies, such as the time to the MRCA and the length of the genealogy, analytically and through efficient simulation. Extensions to the coalescent allow us to incorporate effects of mutation, recombination, selection and demographic events in the coalescent model. A short introduction to the coalescent model can be found here and a longer, more detailed introduction can be read here.

However, coalescent analyses can be slow or suffer from numerical instability, especially for large samples. In a study published earlier this year in Theoretical Population Biology, CEHG fellow Ethan Jewett and CEHG professor Noah Rosenberg proposed fast and accurate approximations to general coalescent formulas and procedures for applying such approximations. Their work also examined the asymptotic behavior of existing coalescent approximations analytically and empirically.

Computational challenges with the coalescent

For a given sample, there are many possible genealogical histories, i.e., tree topologies and branch lengths, which are consistent with the allelic states of the sample. Analyses involving the coalescent therefore often require us to condition on a specific genealogical property and then sum over all possible genealogies that display the property, weighted by the probability of the genealogy. A genealogical property that is often conditioned on is n_t, the number of ancestral lineages in the genealogy at a time t in the past. However, computing the distribution P(n_t) of n_t is computationally expensive for large samples and can suffer from numerical instability.

A general approximation procedure for formulas conditioning on n_t

Coalescent formulas conditioning on n_t typically involve sums of the form f(x)=\sum_{n_t} f(x|n_t) \cdot P(n_t)

For large samples and recent times, these computations have two drawbacks:

–       The range of possible values for n_t may be quite large (especially if multiple populations are being analyzed) and a summation over these values may be computationally expensive.

–       Expressions for P(n_t) are susceptible to round-off errors.

Slatkin (2000) proposed an approximation to the summation in f(x) by a single term f(x|E[n_t]). This deterministic approximation was based on the observation that n_t changes almost deterministically over time, even though it is a stochastic variable in theory. Thus we can write n_t \approx E[n_t]. From Figure 2 in the paper (reproduced here), we can see that this approximation is quite accurate. The authors prove the asymptotic accuracy of this approximation and also prove that under regularity assumptions, f(x|E[n_t]) converges to f(x) uniformly in the limits of t \rightarrow 0 and t \rightarrow \infty . This is an important result since it shows that the general procedure produces a good approximation for both very recent and very ancient history of the sample. Further, the paper shows how this method can be used to approximate quantities that depend on the trajectory of n_t over time, which can be used to calculate interesting quantities such as the expected number of segregating sites in a genealogy.

Approximating E[n_t] for single populations

A difficulty with using the deterministic approximation is that E[n_t] often has no closed-form formula, and if one exists, it is typically not easy to compute when the sample is large.

For a single population with changing size, two deterministic approximations have previously been developed (one by Slatkin and Rannala 1997, Volz et al. 2009 and one by Frost and Volz, 2010, Maruvka et al., 2011). Using theoretical and empirical methods, the authors examine the asymptotic behavior and computational complexity of these approximations and a Gaussian approximation by Griffiths. A summary of their results is in the table below.

Method Accuracy
Griffith’s approximation Accurate for large samples and recent history.
Slatkin and Rannala (1997), Volz et al. (2009) Accurate for recent history and arbitrary sample size, inaccurate for very ancient history.
Frost and Volz (2010), Maruvka et al. (2011) Accurate for both recent and ancient history and for arbitrary sample size.
Jewett and Rosenberg (2014) Accurate for both recent and ancient history and arbitrary sample size, and for multiple populations with migration.

 

Approximating E[n_t] for multiple populations

Existing approaches only work for single populations of changing size and cannot account for migration between multiple populations. Ethan and Noah extend the framework for single populations to allow multiple populations with migration. The result is a system of simultaneous differential equations, one for each population. While it does not allow for analytical solutions except in very special cases, the system can be easily solved numerically for any given demographic scenario.

Significance of this work

The extension of the coalescent framework to multiple populations with migration is an important result for demographic inference. The extended framework with multiple populations allows efficient computation of demographically informative quantities such as the expected number of private alleles in a sample, divergence times between populations.

Ethan and Noah describe a general procedure that can be used to approximate coalescent formulas that involve summing over distributions conditioned on n_t or the trajectory of n_t over time. This procedure is particularly accurate for studying very recent or very ancient genealogical history.

The analysis of existing approximations to E[n_t] show that different approximations have different asymptotic behavior and computational complexities. The choice of which approximation to use is therefore often a tradeoff between the computational complexity of the approximation and the likely behavior of the approximation in the parameter ranges of interest.

Future Directions

As increasingly large genomic samples from populations with complex demographic histories become available for study, exact methods either become intractable or very slow. This work adds to a growing set of approximations to the coalescent and its extensions, joining other methods such as conditional sampling distributions and the sequentially markov coalescent. Ethan and Noah are already exploring applications of these approximate methods to reconciling gene trees with species trees. In the future, I expect that these and other approximations will be important for fast and accurate analysis of large genomic datasets.

References

[1] Jewett, E. M., & Rosenberg, N. A. (2014). Theory and applications of a deterministic approximation to the coalescent model. Theoretical population biology.

[2] Griffiths, R. C. (1984). Asymptotic line-of-descent distributions. Journal of Mathematical Biology21(1), 67-75.

[3] Frost, S. D., & Volz, E. M. (2010). Viral phylodynamics and the search for an ‘effective number of infections’. Philosophical Transactions of the Royal Society B: Biological Sciences365(1548), 1879-1890.

[4] Maruvka, Y. E., Shnerb, N. M., Bar-Yam, Y., & Wakeley, J. (2011). Recovering population parameters from a single gene genealogy: an unbiased estimator of the growth rate. Molecular biology and evolution28(5), 1617-1631.

[5] Slatkin, M., & Rannala, B. (1997). Estimating the age of alleles by use of intraallelic variability. American journal of human genetics60(2), 447.

[6] Slatkin, M. (2000). Allele age and a test for selection on rare alleles.Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences355(1403), 1663-1668.

[7] Volz, E. M., Pond, S. L. K., Ward, M. J., Brown, A. J. L., & Frost, S. D. (2009). Phylodynamics of infectious disease epidemics. Genetics183(4), 1421-1430.

Paper author Ethan Jewett is a PhD student in the lab of Noah Rosenberg.

Paper author Ethan Jewett is a PhD student in the lab of Noah Rosenberg.

Learning from 69 sequenced Y chromosomes

Why the Y?

Blog author Amy Goldberg is a graduate students in Noah Rosenberg's lab.

Blog author Amy Goldberg is a graduate students in Noah Rosenberg’s lab.


While mitochondria have been extensively sequenced for decades because of their short length and abundance, the Y chromosome has been under-studied.  Unlike autosomal DNA, the mitochondria and (most of) the Y chromosome are inherited exclusively maternally and paternally, respectively.  Therefore, they do not undergo meiotic recombination.  Without recombination, mutations accumulate on a stable background, preserving a wealth of information about population history.  Each background, shared through a common ancestor, is called a haplogroup. To leverage this information, Poznik et al. set out to sequence 69 males from nine diverse human populations, including a large representation of African individuals.  The paper, published in Science last summer, is by Stanford graduate student David Poznik and a group lead by CEHG professor Dr. Carlos Bustamante.

The structure of the Y chromosome is complex, with large heterochromatic regions, pseudo-autosomal regions that recombine with the X chromosome, and repetitive elements, making mapping reads difficult.  But, the Y chromosome is haploid, allowing for accurate variant calls at lower coverage than the autosomes, which have heterozygotes.  Using high-throughput sequencing (3.1x mean coverage) and a haploid expectation-maximization algorithm, Poznik et al. called genotypes with an error rate around 0.1%. The paper developed important methods for analyzing high-throughput sequences of the difficult Y chromosome, including determining the subset of regions within which accurate genotypes can be called.

Reconstructing the human Y-chromosome tree

Poznik et al. constructed a phylogenetic tree of the Y chromosome using sequence data and a maximum likelihood approach.  While the overall structure of the tree was known, Poznik et al. were able to accurately calculate branch lengths based on the number of variants differing between individuals and resolve previously indeterminate finer structure.

Figure 2 of the paper: Y-chromosome phylogeny inferred from genomic sequencing.

Figure 2 of the paper: Y-chromosome phylogeny inferred from genomic sequencing.

Incredible African Diversity: One of the key findings of the paper was the depth of diversity within Africans lineages.  While both uniparental and autosomal markers have indicated an African root for human diversity, Poznik et al. find lineages within a single population, the San hunter-gatherers, that coalesce almost at the same time as the entire tree (see haplogroup A). This indicates African diversity and structure has existed for tens of thousands of years, and there is likely more to discover.  A large sample of African populations were considered, which lead to previously unseen structure within haplogroup B2, including structure not mirrored by modern population clustering, that dates to approximately 35,000 years ago.

Evidence of population expansionShort internal branches of the tree, such as those seen within haplogroup E and the non-African group FT, indicate periods of rapid population growth.  When a population expands quickly, new variants that might otherwise drift to extinction can persist.  A large number of coalescence events occur at the time of growth, as there were fewer lineages alive in the population before this time.  For non-African haplogroups, this pattern is likely a remnant of the Out of Africa migration.  For haplogroup E, this corresponds to the Bantu agricultural expansion.

Resolved Eurasian polytomy: Previously, the topology of the Eurasian tree separating haplogroups G-H-IJK was unresolved.  Because of the higher coverage sequencing for this study, Poznik et al. found a single variant, a C to T transition, that differentiates G from the other groups.  Haplogroup G retains the ancestral variant, while H-IJK share the derived variant and are therefore more closely related to each other.

Sequencing vs. genotyping

In contrast to previous studies, which analyzed small repetitive elements called microsatellites or small sets of single base-pair changes called SNPs, whole-genome sequencing data contains not only more information, but potentially more accurate information.  In particular, before the advent of high-throughput sequencing, SNPs were usually ascertained in a subset of individuals that did not capture worldwide diversity levels.  Therefore, diversity measures are often underestimated and biased.  Without sequence data, the branch lengths of the tree did not have a meaningful interpretation, and the depth of variation within Africa was not seen.

MRCA of Human Maternal and Paternal Lineages

There was a lot of public discussion spurred by the publication of Poznik’s paper last year.  The discussion mainly focused on their result that, contrary to previous estimates, the most recent common ancestor (MRCA) of all mitochondrial DNA lived at a similar time as that of all Y chromosomes.  Previous estimates put the mitochondrial TMRCA around 200 thousand years ago, with the Y chromosome coalescing a bit over 100 thousand years ago.  These different estimates for Y and mitochondria were often obtained through different sequencing and analysis methods, and are therefore less comparable.  In particular, varying estimates of the mutation rates have led to different TMRCA estimates.  By analyzing both the Y and mitochondria in the same framework, calibrated by archeological evidence and within-species comparisons, Poznik et al. found largely overlapping confidence intervals for the TMRCA of both Y and mitochondria.

But, should the coalescence times of the mitochondria and the Y chromosome be the same? Not necessarily.  While discrepancies between the mitochondria and Y chromosome have often been interpreted as sex-biased population histories or sizes, strictly neutral models can predict large differences between the two, as well.  Because neither the analyzed part of the Y chromosome nor the mitochondria undergo recombination, each acts as a single locus – and therefore represents the history of a single lineage.  For a population, there is a wide distribution of the ages when lineages would coalesce for a given population history, and these loci represent only two with largely independent histories (given the overall population history), therefore they may differ by chance alone.  Similarly, different loci across autosomal DNA have TMRCA ranging from thousands to millions of years. Additionally, as single loci, any effects of selection would distort the entire genealogy of the Y chromosome and mitochondria.

Future directions

Human population history is far from fully fleshed out, and Poznik et al. provide a framework to leverage increasingly available high-throughput sequencing of Y chromosomes.  The method used to calculate the mutation rate and TMRCA is a valuable contribution in itself, with applications to a wide range of evolutionary and ecological questions.  This study demonstrated that we have only characterized a fraction of worldwide diversity, particularly in Africa, and that increased sampling will be critical to parsing close and far ties in human history.

Reference

Poznik GD, Henn BM, Yee MC, Sliwerska E, Euskirchen GM, Lin AA, Snyder M, Quintana-Murci L, Kidd JM, Underhill PA, Bustamante CD. Sequencing Y chromosomes resolves discrepancy in time to common ancestor of males versus females. Science. 2013 Aug 2;341(6145):562-5. doi: 10.1126/science.1237619.

Paper author David Poznik is a PhD student in Carlos Bustamante's lab.

Paper author David Poznik is a PhD student in Carlos Bustamante’s lab.