8:30AM - Breakfast
9:00AM - Welcome/Introduction [Organizers]
9:10AM - *Keynote Speaker : Gill Bejerano
:: Associate Professor of Developmental Biology, of Computer Science, of Biomedical Data Science and Pediatrics ::
- Title: "Methods, signal everywhere - comparative genomics in the land of plenty"
- Summary: Comparative genomics, an (oft unwanted) sub-field of evo devo, is slowly, if painfully, entering a golden age, with more animal genomes sequenced than you can shake a stick at. I will discuss some fun and extendible methodologies that can help you pick out (and perhaps later CRISPR) beautiful evolutionary signal in most any clade you like. Given time and audience inclination I can also share my thoughts on why it is we have so many genomes, so many methods, and yet so few plug and play tools.
10:30AM - Coffee Break (30min)
11:00AM - 1) Steven Gazal - Price Lab
- Title: Low-frequency variant architecture of human complex traits reveal strength of negative selection across coding and non-coding annotations
- Summary: Low-frequency variants (0.5% ≤ MAF < 5%) can have larger per-allele effect sizes than common variants (MAF ≥ 5%) when impacted by negative selection, and may thus yield important biological insights on the genetic architecture of human complex traits (even though the heritability they explain is modest). However, little is known on how low-frequency variant heritability is distributed across functional annotations. Here, we partitioned the heritability of both low-frequency and common variants in 40 UK Biobank traits across a broad set of functional annotations. We highlighted the critical impact of low- frequency non-synonymous variants compared to common non-synonymous variants, as well as low- frequency variants in annotations previously linked to negative selection. Non-coding annotations that were significantly enriched for common variants tended to be similarly enriched for low-frequency variants, but more enriched for brain-related annotations and traits. Forward simulations confirmed that low-frequency variants inform on the mean selection coefficient of causal variants in the annotation.
11:25AM - 2) Maria Alexis - Burge Lab
- Title: Conservation of RNA regulatory elements
- Summary: RNA-binding proteins (RBPs) are highly conserved factors that regulate the production and processing of mRNAs, by binding 3-8 nucleotide motifs in their targets. How RNA regulatory motifs may have co-evolved with their cognate RBPs, the impact of mutations on RBP binding patterns, and the extent to which regulation is conserved remain poorly understood. Recent efforts aimed at cataloguing RBP specificities have determined the spectrum of affinities to all possible sequence motifs for over 100 RBPs. Using these data, we observed a strong positive correlation between strength of binding and motif conservation in the transcriptome, suggesting co-evolution between trans factor (RBP) specificities and the cis sequence elements they bind. Across species, changes in RBP binding sites are enriched for mutations that preserve binding, however, many mutations result in preferential binding of a different RBP, indicating RNA regulation is rarely lost but also highly evolvable between different factors. Comparing in vivo binding profiles of a candidate RBP (UNK) in human and mouse neuronal cells, we find that many target genes are bound in both species. Interestingly, while some UNK binding sites are conserved, a large fraction binding sites are species-specific, suggesting a pressure to conserve motif frequencies rather than specific binding sites. In support of this hypothesis, we also show that genes with poor sequence alignments that have similar motif composition are more similarly regulated across species. These results demonstrate that RNA regulation is highly conserved and that conservation can often be underestimated by traditional alignment- based metrics. This work highlights the use of in vitro RBP binding assays to better understand evolutionary patterns of the transcriptome.
11:50AM - Lunch
1:10PM - 3) Nate Edelman - Mallet Lab
- Title: Hybridization throughout the Heliconius radiation
- Summary: Through de novo genome assembly and alignment of 21 species, we reveal an evolutionary history of admixture that characterizes the Heliconius butterfly radiation. Hybridization is observed throughout the tree of life, and amongst eukaryotes is especially prevalent during rapid radiations. As they diverge, many closely related species overlap in both time and space, providing ample opportunities for interspecific gene sharing. Most studies of hybridization during radiations rely on population-level data, and aim to identify present-day hybridization with unclear implications over macro-evolutionary time scales. By contrast, the Heliconius radiation provides an opportunity for us to observe the genomic signatures of admixture over 12 million years of evolution. Here, we use DISCOVAR de novo to produce whole genome assemblies for 21 species in the Heliconiini sub-family, and generate a multiple sequence alignment with progressiveCactus. We then use both allele frequency and gene tree based methods to infer phylogenetic networks for our subset of Heliconius species. We find that admixture has been common throughout the radiation. In one of the two sub-clades we identify at least two major introgression events, while the other contains too much admixture to be confidently resolved by present methods. We also identify two large inversions that exemplify incomplete lineage sorting and introgression, and could represent modules of genes selected as a block. The abundance of non-treelike relationships adds emphatically to the growing body of evidence that the process of evolution is better represented as a network than a bifurcating tree.
1:35PM - 4) Michael Miyagi - Wakeley/Desai groups
- Title: Inferring Properties of Introgression Events Using Gene Tree Branch Lengths
- Summary: Gene tree and species tree discordance can either be indicative of reticulate evolution, such as introgression, or be a result of deep coalescence events (also known as incomplete lineage sorting). In order to differentiate between these signals, we use coalescent theory to describe the behavior of the internal branch length of three-leaf gene trees generated from a species tree with at most one introgression event. Using this model, we demonstrate that this internal branch length distribution is sufficient for the diagnosis of the presence, direction, and timing of introgression. We implement this result in a procedure for arbitrarily large species trees which does not require a priori knowledge of putative hybrid populations, and use this to quantify reticulate evolution in the Heliconius phylogeny.
2:00PM - Coffee Break (30min)
2:30PM - 5) Andreas Kautt - Hoekstra Lab
- Title: Genomics of adaptation and sympatric divergence in Neotropical crater lake cichlid fish
- Summary: That speciation may happen in the face of gene flow is now widely accepted. Yet, empirical studies elucidating when and how selection can overcome gene flow and lead to genome- wide differentiation are still few. Midas cichlids in Nicaragua are a powerful natural system to investigate the conditions conducive to speciation-with-gene-flow, since these fish have repeatedly colonized isolated crater lakes and speciated in some of them but not others. Here, in combination with a recently generated state-of-the-art reference genome, we re-sequenced the genomes of more than 450 individuals to characterize the genomic architecture of ecologically-relevant traits and the genome-wide landscape of speciation.
2:55PM - 6) Evelyn Jagoda - Capellini Lab
- Title: Investigations into the Phenotypic Consequences of Archaic Hominin Introgression in Living Humans
- Summary: A fundamental question in human biology is, what are the adaptive genetic variants that underlie the remarkable ability of humans to inhabit diverse environments? While this is a daunting question in part due to the sheer size and diversity present in modern genomes, studies comparing ancient and modern genomes have added a new dimension to this question. They have revealed long stretches of DNA sequences (haplotypes), distributed across modern genomes that were introduced by interbreeding with archaic hominins. A subset of these haplotypes are under positive selection and have played beneficial roles during human dispersal and adaptation to different environments across the globe. Yet, as in all genomic studies, narrowing down these introgressed haplotypes to the actual causative base-pair(s) remains an enormous task. We are using a variety of computational and wet laboratory approaches track down these to identify actual causative base-pair(s) and their specific phenotypic consequences. By helping to understand phenotypes that were adaptive in the past and the genomic architecture of these phenotypes today, this research will provide a lens through which to view human adaptation to a the modern environment of today.
3:20PM - Closing Remarks [Organizers]