We assembled a chromosome-scale genome. To our knowledge, this is the most contiguous coral genome assembled to date - scaffold N50>19Mb (2/n)
-
-
Prikaži ovu nit
-
Led by
@LineKBay we collected over 250 samples in the wild from the central GBR during peak bleaching in 2017 from 12 reefs. We resequenced 44 of these genomes at high coverage and built a reference imputation panel from 44 samples (3/n)Prikaži ovu nit -
With
@joe_pickrell and@Gencove, we imputed genotypes for nearly 200 samples we sequenced at low-coverage (w/@pandolfatto). We show we can impute common variants with high accuracy (4/n)Prikaži ovu nit -
From our whole-genome sequencing approach, we also show that we can determine the symbiont type and can use the relative abundance of symbiont reads as a quantitative measure of bleaching (5/n)
Prikaži ovu nit -
What did we learn from these data? First, we look at population structure and demographic history (6/n)
Prikaži ovu nit -
We find minimal population structure and high connectivity across the sampled reefs. All individuals show very similar long-term demographic histories, with a steady decline in Ne over > 1 Mya to 15 Kya (7/n)
Prikaži ovu nit -
With high gene flow across a heterogenous environment, models suggest adaptively evolving loci should show a form of balancing selection, so we searched for regions of elevated diversity (8/n)
Prikaži ovu nit -
One gene in particular stands out with two common, highly diverged haplotypes - sacsin, a heat-shock co-chaperone (9/n)
Prikaži ovu nit -
We confirmed these findings were not a bioinformatic artifact by resequencing this specific region. We show the gene has been under long term balancing selection and has an unusually deep genealogy (10/n)
Prikaži ovu nit -
We then turn our attention to a GWAS for variation in bleaching and to the development of a polygenic score and predictive model (11/n)
Prikaži ovu nit -
No single SNP reach genome-wide significance, indicating the absence of common, large effect loci underlying variation in bleaching response (12/n)
Prikaži ovu nit -
But prior evidence strongly points to bleaching response being partly heritable in the coral, so are the combined estimated effects of variants genome-wide predictive of bleaching? (13/n)
Prikaži ovu nit -
Through cross-validation, we show that this answer is yes (but a weak yes). While a polygenic score for bleaching gives a significant increase in prediction accuracy, it is a small increase on its own (14/n)
Prikaži ovu nit -
However, combining a polygenic score with environmental data and information for the dominant symbiont species, we can explain >60% of the variation in bleaching for our samples (using cross-validation) (15/n)
Prikaži ovu nit -
We view this work as an important first step towards incorporating genomics to study traits of conservation importance. While still a massive dataset, ~200 individuals for a GWAS leaves us under powered in the absence of major effect loci (16/n)
Prikaži ovu nit -
As sample sizes increase, so will our prediction accuracy and power to detect associations. Here, we show that such an approach is feasible in A. millepora and we are already taking the next steps towards this end (17/n)
Prikaži ovu nit -
Supported by the Agouron Institute, new collections and field work are underway as we speak for
#1KCoralGenomes led by@LineKBay and@SeaSim_AIMS (18/n)Prikaži ovu nit -
We released all the data and provide an extensive supplement that we hope can be helpful for similar genomics based approaches in other non-model organisms (19/n)
Prikaži ovu nit -
This was a large undertaking involving many people, so there are many thanks to go around - first of course to
@molly_przew (20/n)Prikaži ovu nit -
All co-authors had major contributions to this project, so a sincere thank you to all involved! Also many helpful conversations with other along the way, including
@MollySchumer and@arbelharpak, and@richard_durbin for the assembly (21/n)Prikaži ovu nit -
We are looking forward to any comments or feedback, please contact me with any questions. Excited for this work to continue in the future and where the next findings take us (23/n)
Prikaži ovu nit
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.