A Nice opinion on confronting uncertainty and modeling it for GBS data

Just over a week ago, I had the opportunity to work in Chris Nice‘s lab at Texas State University. I was accompanied by one of our MS students, Ben, and my colleague, Erik Sotka, to prep libraries for a genomic survey of a certain alga I’ve a penchant to write about. We also were there to prep a library with Torrance Hanley, a postdoc in the Kimbro and Hughes labs at Northeastern.

Chris walked us through each step as we embark on our first population genomic projects. We got to talking about analyses and issues I’ve written about before. In addition, we got to talking about times in which Bayesian approaches, such as STRUCTURE, may not be appropriate (i.e., when there are strong departures from HWE) and possible ways to get around this in the future!

I asked Chris to offer his opinion and write a small piece for TME. Et voilà 

Population genomics is certainly progressing as a field and there seems to be about as many ways to do things as there are labs doing them. Several methods for library construction have been reviewed recently with some good discussions (Andrews & Luikart, 2014; Puritz et al., 2014; Andrews et al., 2014). One area that has not received as much attention is the downstream analytical details – once you have your sequence reads.

Continue reading

Posted in bioinformatics, genomics, interview | Tagged , , , | 1 Comment

Conservation through hybridization

Mus spretus Photo credit

The Algerian mouse Mus spretus “lent” its pesticide resistance alleles to the European House mouse. How kind! Photo credit University of Valencia 


 
 
 
 
 
 
 
 
 
 
 
Human actions have contributed to an unprecedented increase in atmospheric CO2 in the last 250 years, and global temperatures will increase by as many as 4°C by the end of the century. Species can respond to changing conditions in three ways:                     i) individuals and populations can shift their ranges and migrate to more hospitable habitats, ii) individuals can acclimatize to current conditions through physiological plasticity, or iii) populations can adapt through evolution. If none of these options are possible, the species is faced with local or global extinction.
The capacity of a species or population to adapt to climate change depends in part on standing genetic variation, population size, and mutation rate. In an essay recently published in Conservation Biology, Jill Hamilton and Joshua Miller (2015) discuss the potential for gene flow between different species to act as a kind of rescue for species that have low adaptive capacity, injecting beneficial alleles into vulnerable populations, thus facilitating their persistence in the face of climate change. Continue reading

Posted in Uncategorized | Leave a comment

To the final estuaries

For the final stop on our Japanese sampling leg, we ventured to the most populous metropolitan area in the world.
Tokyo was known as Edo (江戸), or estuary, until it became the imperial capital in 1868. An apt location to end our field expedition to many estuarine populations of Gracilaria vermiculophylla in Japan.
Continue reading

Posted in blogging, domestication, evolution, haploid-diploid | Tagged , , , | Leave a comment

Genomics of domestication in chicken and cattle

Two recent studies attempt to understand the process of adaptive evolution in domestication and artificial selection by characterizing (a) sweeps, and their association with phenotypes in extant hybrid lines (Sheng et al. 2015), and (b) phylogenomic position of an extinct wild ancestor (Park et al. 2015).
Sheng et al. (2015) – Selection in Virginia chicken lines

HWS and LWS lines from the Virginia chicken experiment. Image courtesy: http://www.vt.edu/spotlight/innovation/2011-03-07-poultry/poultry.html


Artificial selection experiments provide a classic means to study the genomic consequences of accelerated adaptive evolution, and differentiating between standing genetic variation and novel polymorphisms that contribute to the evolution of advantageous phenotypes. Sheng et al. (2015) estimate the genome-wide effects of long-term bi-directional selection on 56-day body weight in experimental populations of Virginia chicken lines, derived from a common ancestral population, and characterize sweeps contributing to adaptive evolution. Specifically, by genotyping (at 106 sweep loci, previously identified to be fixed for alternate alleles) and phenotyping a large F15 line (Advanced Intercrossed Line – AIL) of hybrids from body weight variant lines, Sheng et al. (2015) report (1) 99 sweeps to be potentially adaptive, (2) most identified alternate alleles at sweeping loci were associated with increased weight, (3) significant association of the length of sweeping loci with strength of selection, (4) sweeping alleles were likely present in the ancestor, thus stressing the role of standing genetic variation in adaptive evolution.

A large number of loci, each having small allele-substitution effects, were major contributors to the selectable additive variance during adaptation. The loci that were fixed in the divergent lines after 40 generations of selection contributed much of the variation in the base population, thus providing empirical support for earlier work that has suggested that initial selection response is likely to result from selection on standing genetic variation…In summary, these results provide not only novel insights to the genetics contributing to the gradual, continued, long-term response to selection in the Virginia lines, but also to the fundamental genetic mechanisms contributing to selection and adaptation.

Park et al. 2015 – Evolutionary history of wild aurochs

Extinct wild aurochs (Bos primigenius) skeleton. Image courtesy: http://www.pbase.com/bmcmorrow/image/144858805


Domestication of the now extinct wild auroch, Bos primigenius gave rise to two the now extant cattle species, Bos taurus and Bos indicus, previously reported to have occurred in the Near East and Southwest Asia. Park et al. (2015) sequence the genome of B. primigenius from bone powders extracted from six well preserved aurochs bones of the CPC98 sample. Upon alignment the B. taurus genome, the CPC98 sample was determined to be from a male, and > 5 million variants were identified, with > 2 million SNP’s passing quality filters. A maximum likelihood phylogenetic tree constructed using ~15k SNP’s across the CPC98 sample, and 1228 extant B. taurus and B. indicus cattle supports separate domestic origins of the two extant species, and the CPC98 sample as an outgroup to the modern B. taurus clade, also corroborated by a PCA. Analyses of population structure however reveal the presence of three putative subpopulations (European, African, and zebu) in the CPC98 sample, indicating that several of the allelic variants now private to extant clades may have been present more widely prior to domestication. ABBA-BABA tests of genomic admixture between aurochs and extant European taurine cattle also provide significant support for gene flow from the now extinct B. primigenius into European cattle, also shown by their TreeMix analyses. HKA tests of selection in the B. taurus lineage identified several regions under positive selection, and characterized several functional elements involved in neurobiology, muscle development and function, and immune response genes.

First 2 PC’s of variation using SNP data from B. taurus, B. indicus, and B. primigenius lines. Image courtesy: Park et al. (2015)

Sequencing data from the aurochs genome provides an important reference for testing hypotheses regarding the genetic consequences of recent artificial and natural selection in modern cattle. In addition, future analyses of whole-genome sequences from modern animals and additional aurochs samples will refine the catalogue of genomic loci that contribute to key functional traits in domestic cattle.

 
References
Sheng, Z, et al. (2015) Standing genetic variation as a major contributor to adaptation in the Virginia chicken lines selection experiment. Genome Biology. DOI: 10.1186/s13059-015-0785-z
Park, S, et al. (2015) Genome sequencing of the extinct Eurasian wild aurochs, Bos primigenius, illuminates the phylogeography and evolution of cattle. Genome Biology. DOI: 10.1186/s13059-015-0790-2
 

Posted in adaptation, bioinformatics, domestication, evolution, genomics, natural history, Paleogenomics, phylogeography, population genetics, selection, speciation, STRUCTURE | Tagged , , , , , , | 1 Comment

Three Views of Japan

By the time we reached Sendai, we were heading into our fourth week of sharing one tiny suitcase of clothes, while bags of silica were luxuriously spread across three large suitcases! Games of Jenga in the teeny rental cars were losing their initial appeal and humor!
We’d seen varied land- and seascapes, but our third stop was home to one of the Three View of Japan, Matsushima Bay. There are over 250 pine (matsu) covered islands (shima).

Matsushima Bay

Field site looking out toward Matsushima Bay


Despite the proximity of these picturesque islands to the devastation wrought by the 2011 earthquake and tsunami, Matsushima was largely protected.
This was not the case for the rest of the northern Honshu coastline, particularly in the Miygai and Iwate Prefectures.
Energy map of the tsunami from NOAA

Energy map of the tsunami. The dark red is centered along the Miyagi and Iwate Prefectures © NOAA


The earthquake was the fourth most powerful recorded, worldwide, since records have been kept and even shifted the Earth on its axis. New York Times article following the catastrophe described how Japan jumped about 4 m (13 feet) closer to North America. In addition, a 400 km (250 mi) stretch of the Japanese coastline dropped in altitude by half a meter (2 feet), facilitating the tsunami to travel farther and faster over the coast inland.
The tsunami reached over 40 m (greater than 130 feet!) in the Iwate Prefecture.! In Sendai, where Tohoku University and our hosts are located, water traveled up to 10 km (6 mi) inland.
This combination of three photos taken over a six month period shows the March 11 tsunami and its aftermath at Sendai Airport in Sendai, Miyagi prefecture, northern Japan. The top photo taken March 11, 2011 shows the tsunami engulfing the airport immediately after an earthquake. The middle photo, taken June 3, 2011 and the bottom photo, taken Sept. 6, 2011 show the restored and reopened airport. © AP / Kyodo News

This combination of three photos taken over a six month period shows the March 11 tsunami and its aftermath at Sendai Airport in Sendai, Miyagi prefecture, northern Japan. The top photo taken March 11, 2011 shows the tsunami engulfing the airport immediately after an earthquake. The middle photo, taken June 3, 2011 and the bottom photo, taken Sept. 6, 2011 show the restored and reopened airport. © AP / Kyodo News


The parking lot of the Sendai airport immediately following the tsunami. © Roberto De Vido/TucsonSentinel.com

The parking lot of the Sendai airport immediately following the tsunami. © Roberto De Vido/TucsonSentinel.com


Sendai Airport car park right near above image. Note sign showing the water level. These are throughout the region now.

Sendai Airport car park. This image was taken in 2015 a bit to the left from where the image above, by Roberto De Vido, was taken.


A train station near Sendai. Rob is picture beneath one of the blue signs documenting water height. He is 1.93 m (6'4") and the sign read over 3 m of standing water.

A train station near Sendai. Rob is pictured beneath one of the blue signs documenting the height of the water. He is 1.93 m (6’4″) and the sign read over 3 m (almost 10 feet!) of standing water.


The earthquake and tsunami clean-up continues to this day. The drives we took around the area were sobering. Foundations are all that is left of once thriving towns. Eerily, the GPS in our rental car (we thought we should rent Lamborghini), still showed where hotels and restaurants should have been.

Continue reading

Posted in blogging, community, evolution, haploid-diploid, natural history | Tagged , , , , , | Leave a comment

Bigger, badder males drive unbalanced genetic introgression in wall lizards


Female mate choice is one of the major barriers to hybridization. But even when hybridization does happen, females are often identified as the primary drivers of how phenotypes move through a hybrid zone. A new paper appearing in Ecology Letters by While et al. goes against this trend by showing that male dominance drives asymmetric introgression in common wall lizards (Podarcis muralis).
In a very cool and integrative paper, While et al. established phenotypes (morphology, coloration, bite force and testes mass) likely under sexual selection in the core ranges of two wall lizard lineages (one from western Europe and one from northern Italy). They then created experimental populations that mimic a hybrid zone to describe male-male interactions and any asymmetries in reproductive success, including the survival of F1 hybrids. To make sure that any differences in mating success by the males of each lineage weren’t due to some post-copulatory mechanism, they additionally conducted sperm competition trials.
The Italian lineage males were all-around more competitive (larger heads, stronger bites, greater testes mass) and these phenotypes lead to dominance over males from the Western European lineage in the experimental populations: females from the Western European lineage were much more likely to give birth to young fathered by the opposite lineage.
The authors then looked at genetic patterns of introgression at three locations (one natural hybrid zone and two introduced areas) to understand if the results of their experiments match the observed patterns of genetic introgression in the wild. As predicted, the natural area of hybridization between the two lineages in northwestern Italy shows a much greater westward cline shift for microsatellites compared to that by mitochondrial DNA.

Our results show that divergence in male competitive ability in allopatry causes asymmetric hybridisation and gene flow upon secondary contact in wall lizards. As a consequence, sexually selected introgression shapes phenotypic and genetic variation in both native and non-native populations.

Male-male conflict may be a driver of genetic introgression in systems only when female choice is relatively weak (like in lizards), but this paper provides 1) an interesting perspective on how asymmetries in introgression are perpetuated and 2) a reason to investigate this phenomenon in other vertebrates that have varying strengths of female mate-choice.
 
While, G. M., Michaelides, S., Heathcote, R. J., MacGregor, H. E., Zajac, N., Beninde, J., … & Uller, T. (2015). Sexual selection drives asymmetric introgression in wall lizards. Ecology Letters. DOI: 10.1111/ele.12531

Posted in Uncategorized | Leave a comment

On false positives in Isolation with Migration analyses

The IM suite of tools (IM, IMa, IMa2, IMa2p, etc.) are used widely by molecular ecologists at large for the analyses and estimation of ancestral demography under an Isolation with Migration (IM) model. However, these tools come with fundamental assumptions on the evolutionary processes underlying genomic loci under question – loci are assumed to be independent, freely recombining between loci, non-recombining within loci, and putatively neutral. These methods have been previously found to be robust for inference using data-sets that largely fit these assumptions. In a recent publication, we (Hey et al. 2015) analyzed via simulations, conditions under which utilizing likelihood ratio tests for inference under an Isolation with migration (IM) model may result in an excess of false positives for the presence of migration – an observation previously reported by Cruickshank and Hahn (2014). Particularly, we were interested in the case of data sets we call “SDLD”, or “Small Data Low Divergence” – i.e. the number of loci sampled is small (< 5 as reported by Cruickshank and Hahn (2014)), and exhibiting very low divergence between populations.
We simulated 20 data sets with two loci, low divergence (t=0.5), no migration (m=0), and analyzed these data using a modified version of IMa2, which approximates the joint posterior density of migration rates, and divergence times (having integrated out the population size parameters). Our key findings include: (1) high false positive rates for migration, (2) joint posterior density estimates indicate areas of high posterior densities for models with low m and t, and models with high m, and t, indicating model identifiability issues in the SDLD context, (3) these very different models show similar expected allele frequency spectra, and differentiation distributions (measured as Weir and Cockerham’s φst).

Two population AFS under two simulated scenarios - Fig. A showing  the case of low m, and low t, versus Fig. B showing the case of high m, and high t. Image courtesy: Hey et al. (2015) DOI: 10.1111/mec.13381

Two population AFS under two simulated scenarios – Fig. A showing the case of low m, and low t, versus Fig. B showing the case of high m, and high t. Image courtesy: Hey et al. (2015) DOI: 10.1111/mec.13381


 
Besides cautioning against poor sampling (an issue which I have also previously discussed here), our study also points to a high false positive rates for detecting migration using likelihood ratio tests while using the IM suite of tools on data that show low divergence (eg. very low Fst), and while using a small number of loci.
References:
Hey, J., Chung, Y, & Sethuraman, A. (2015). “On the occurrence of false positives in tests of migration under an isolation-with-migration model.” Molecular Ecology DOI: 10.1111/mec.13381
Cruickshank, Tami E., and Matthew W. Hahn. “Reanalysis suggests that genomic islands of speciation are due to reduced diversity, not reduced gene flow.” Molecular Ecology 23.13 (2014): 3133-3157.

Posted in evolution, genomics, howto, IMa2, methods, Molecular Ecology views, natural history, population genetics, software, theory | Tagged , , , , , , , | 1 Comment

Along the Mackerel Road

We left Hokkaido and flew to Osaka where we collected our next rental car (our first teeny tiny one!)

With A LOT of field gear, tiny cars in Japan and Europe can pose interesting spatial problems!

With A LOT of field gear, tiny cars in Japan and Europe can pose interesting spatial problems!


As our flight was delayed due to weather in Hokkaido, we decided to break the journey between Osaka and Obama (it means “little ocean” in Japanese) in Kyoto.
Fushimi Inari-taisha in Kyoto

Fushimi Inari-taisha in Kyoto


D7000s_3385_bell kyoto
The Ninomaru Palace of the Nijo-jo, the shogun's residence! The "nightingale floors" (uguisubari) in the corridors protected the occupants from sneak attacks and assassins. The floors squeak as you walk and sound like nightingales!

The Ninomaru Palace of the Nijo-jo, the shogun’s residence!
The “nightingale floors” (uguisubari) in the corridors protected the occupants from sneak attacks and assassins. The floors squeak as you walk and sound like nightingales!


Cars with character

Cars with character


Continue reading

Posted in adaptation, blogging, community, haploid-diploid, natural history, population genetics, speciation | Tagged , , , , , , | Leave a comment

Genetic distance predicts the spread of deadly fungal infections in bats

Little brown bat (Myotis lucifugus)


You’ve probably heard about White-Nose Syndrome (WNS), the particularly nasty fungal pathogen that has decimated North American bat populations over the last decade. Not only has WNS been extremely deadly, but the speed at which it’s spread has been alarming. Really alarming:

Yikes. Original map and more information here.


Understanding why WNS has spread so quickly is an obvious step towards predicting the future spread and potential control of the disease, but it is not easy task. Mainly, the spread of WNS hasn’t radiated out linearly from a starting point. Instead, outbreaks have hopscotched across eastern North America, often showing up in disjunct and distant populations from year to year:

For example, WNS appeared as far south as southern Virginia before it reached the western edge of New York. Previous analyses indicate that colony size, species composition, habitat patchiness and climate influence WNS risk and consequently the overall pattern of spread across the landscape (Maher et al. 2012; Thogmartin et al. 2012; Wilder et al. 2011).

But the movement of pathogens from population to population depends on dispersal of their host, right? While straight line distance might not be the best predictor for the spread of WNS, undetected barriers for bat dispersal might play a large role. A recently-accepted paper in Molecular Ecology by Aryn Wilder et al. (quoted above) provides a range-wide analysis of genetic structure of little brown bats (Myotis lucifugus) that shows how inferences of recent dispersal informs the spread of WNS.
Wilder and colleagues collected mtDNA and RADseq data from individuals that represented the entire North American range of little brown bats. The resultant genetic distances between populations were considered along with geographic distances between sites in a mixed-effects model that described the real-life spread of WNS across years. Predictably, the more distantly-related or distantly-located two populations were, the less likely their risk WNS introduction. However, the the highest level of predictability was demonstrated when genetic covariates are included in the model, showing that distance by itself had lower predictive power for the spread of WNS.
The impressive long-distance dispersal by little brown bats (regularly > 500 km; Norquay et al. 2013) and the paltry genetic differentiation across all sites measured with RAD loci (ΦST = 0.008) combines to paint a picture of WNS spreading through significant bat dispersal. However, the lack genetic structure seen in eastern North America (where WNS has spread the most) doesn’t carry over to populations in the west:

Our range-wide genomic data suggest that the spread of WNS by an important host is likely to be slower in western North America than in the eastern half of the continent, where a rapid geographic expansion of the disease has been observed over the past eight years. Given that the spread of WNS to date has been correlated with subtle patterns of genetic differentiation in eastern North America, the rate of dispersal of the fungal pathogen by little brown myotis should decrease as WNS moves into the Great Plains, where genetic differentiation among localities increases substantially.

 
Wilder, A. P., Kunz, T. H., & Sorenson, M. D. (2015). Population genetic structure of a common host predicts the spread of white‐nose syndrome, an emerging infectious disease in bats. Molecular Ecology. DOI: 10.1111/mec.13396

Posted in Molecular Ecology, the journal, phylogeography, population genetics | Tagged , | Leave a comment

Machine learning for model selection in population genomics

Machine learning. Image from www.toptal.com

Machine learning. Image from www.toptal.com


The application of model-based methods in phylogeography helped the field transition from a more qualitative, overlay-a-tree-on-a-map, discipline to one that tests hypotheses in robust statistical frameworks. Many researchers have embraced approximate Bayesian computation (ABC) for model selection since computing the likelihood of complex population models is often intractable. While ABC is easy to use and provides a posterior distribution, like all methods, it has its weaknesses. For example, ABC requires a very large amount of simulated data and the choice of what summary statistic and rejection threshold to use is difficult.
In a recent preprint posted on bioRxiv, Sara Sheehan and Yun Song present a likelihood-free inference framework for population genomics that applies deep learning, an active area of machine learning research. They aimed to jointly infer natural selection and changes in population size, processes that can leave similar signatures in the genome, by testing their method on simulated data and on empirical data for Drosophila melanogaster. Sheehan and Song provide accessible introductions to ABC and deep learning in their preprint that are well worth the read. Continue reading

Posted in Uncategorized | Leave a comment