Understanding the consequences of local adaptation at the genomic diversity is a central goal in evolutionary genetics of natural populations. In species with large continuous geographical distributions the phenotypic signal of local adaptation is frequently clear, but the genetic basis often remains elusive. We examined the patterns of genetic diversity in Pinus sylvestris, a keystone species in many Eurasian ecosystems with a huge distribution range and decades of forestry research showing that it is locally adapted to the vast range of environmental conditions. Making P. sylvestris an even more attractive subject of local adaptation study, population structure has been shown to be weak previously and in this study. However, little is known about the molecular genetic basis of adaptation, as the massive size of gymnosperm genomes has prevented large scale genomic surveys. We generated a both geographically and genomically extensive dataset using a targeted sequencing approach. By applying divergence-based and landscape genomics methods we identified several loci contributing to local adaptation, but only few with large allele frequency changes across latitude. We also discovered a very large (ca. 300 Mbp) putative inversion potentially under selection, which to our knowledge is the first such discovery in conifers. Our results call for more detailed analysis of structural variation in relation to genomic basis of local adaptation, emphasize the lack of large effect loci contributing to local adaptation in the coding regions and thus point out the need for more attention towards multi-locus analysis of polygenic adaptation.

The high climatic variability in the past hundred thousand years has affected the demographic and adaptive processes in many species, especially in boreal and temperate regions undergoing glacial cycles. This has also influenced the patterns of genome-wide nucleotide variation, but the details of these effects are largely unknown. Here we study the patterns of genome-wide variation to infer colonization history and patterns of selection of the perennial herb species Arabidopsis lyrata, in locally adapted populations from different parts of its distribution range (Germany, UK, Norway, Sweden, and USA) representing different environmental conditions. Using site frequency spectra based demographic modeling, we found strong reduction in the effective population size of the species in general within the past 100,000 years, with more pronounced effects in the colonizing populations. We further found that the northwestern European A. lyrata populations (UK and Scandinavian) are more closely related to each other than with the Central European populations, and coalescent based population split modeling suggests that western European and Scandinavian populations became isolated relatively recently after the glacial retreat. We also highlighted loci showing evidence for local selection associated with the Scandinavian colonization. The results presented here give new insights into postglacial Scandinavian colonization history and its genome-wide effects.

STUDY QUESTION Can we identify novel variants associated with polycystic ovary syndrome (PCOS) by leveraging the unique population history of Northern Europe? SUMMARY ANSWER We identified three novel genome-wide significant associations with PCOS, with two putative independent causal variants in the checkpoint kinase 2 (CHEK2) gene and a third in myosin X (MYO10). WHAT IS KNOWN ALREADY PCOS is a common, complex disorder with unknown aetiology. While previous genome-wide association studies (GWAS) have mapped several loci associated with PCOS, the analysis of populations with unique population history and genetic makeup has the potential to uncover new low-frequency variants with larger effects. STUDY DESIGN, SIZE, DURATION A population-based case–control GWAS was carried out. PARTICIPANTS/MATERIALS, SETTING, METHODS We identified PCOS cases from national registers by ICD codes (ICD-10 E28.2, ICD-9 256.4, or ICD-8 256.90), and all remaining women were considered controls. We then conducted a three-stage case–control GWAS: in the discovery phase, we had a total of 797 cases and 140 558 controls from the FinnGen study. For validation, we used an independent dataset from the Estonian Biobank, including 2812 cases and 89 230 controls. Finally, we performed a joint meta-analysis of 3609 cases and 229 788 controls from both cohorts. Additionally, we reran the association analyses including BMI as a covariate, with 2169 cases and 160 321 controls from both cohorts. MAIN RESULTS AND THE ROLE OF CHANCE Two out of the three novel genome-wide significant variants associating with PCOS, rs145598156 (P = 3.6×10−8, odds ratio (OR) = 3.01 [2.02–4.50] minor allele frequency (MAF) = 0.005) and rs182075939 (P = 1.9×10−16, OR = 1.69 [1.49–1.91], MAF = 0.04), were found to be enriched in the Finnish and Estonian populations and are tightly linked to a deletion c.1100delC (r2 = 0.95) and a missense I157T (r2 = 0.83) in CHEK2. The third novel association is a common variant near MYO10 (rs9312937, P = 1.7 × 10−8, OR = 1.16 [1.10–1.23], MAF = 0.44). We also replicated four previous reported associations near the genes Erb-B2 Receptor Tyrosine Kinase 4 (ERBB4), DENN Domain Containing 1A (DENND1A), FSH Subunit Beta (FSHB) and Zinc Finger And BTB Domain Containing 16 (ZBTB16). When adding BMI as a covariate only one of the novel variants remained genome-wide significant in the meta-analysis (the EstBB lead signal in CHEK2 rs182075939, P = 1.9×10−16, OR = 1.74 [1.5–2.01]) possibly owing to reduced sample size. LARGE SCALE DATA The age- and BMI-adjusted GWAS meta-analysis summary statistics are available for download from the GWAS Catalog with accession numbers GCST90044902 and GCST90044903. LIMITATIONS, REASONS FOR CAUTION The main limitation was the low prevalence of PCOS in registers; however, the ones with the diagnosis most likely represent the most severe cases. Also, BMI data were not available for all (63% for FinnGen, 76% for EstBB), and the biobank setting limited the accessibility of PCOS phenotypes and laboratory values. WIDER IMPLICATIONS OF THE FINDINGS This study encourages the use of isolated populations to perform genetic association studies for the identification of rare variants contributing to the genetic landscape of complex diseases such as PCOS. STUDY FUNDING/COMPETING INTEREST(S) This work has received funding from the European Union’s Horizon 2020 research and innovation programme under the MATER Marie Skłodowska-Curie grant agreement No. 813707 (N.P.-G., T.L., T.P.), the Estonian Research Council grant (PRG687, T.L.), the Academy of Finland grants 315921 (T.P.), 321763 (T.P.), 297338 (J.K.), 307247 (J.K.), 344695 (H.L.), Novo Nordisk Foundation grant NNF17OC0026062 (J.K.), the Sigrid Juselius Foundation project grants (T.L., J.K., T.P.), Finska Läkaresällskapet (H.L.) and Jane and Aatos Erkko Foundation (H.L.). The funders had no role in study design, data collection and analysis, publishing or preparation of the manuscript. The authors declare no conflicts of interest.

Understanding the consequences of local adaptation at the genomic diversity is a central goal in evolutionary genetics of natural populations. In species with large continuous geographical distributions the phenotypic signal of local adaptation is frequently clear, but the genetic background often remains elusive. We examined the patterns of genetic diversity in Pinus sylvestris, a keystone species in many Eurasian ecosystems with a huge distribution range and decades of forestry research showing that it is locally adapted to the vast range of environmental conditions. Making P. sylvestris an even more attractive subject of local adaptation study, population structure has been shown to be weak previously and in this study. However, little is known about the molecular genetic basis of adaptation, as the massive size of gymnosperm genomes has prevented large scale genomic surveys. We generated a both geographically and genomically extensive dataset using a targeted sequencing approach. By applying divergence-based and landscape genomics methods we found that several coding loci contribute to local adaptation. We also discovered a very large (ca. 300 Mbp) putative inversion with a signal of local adaptation, which to our knowledge is the first such discovery in conifers. Our results call for more detailed analysis of structural variation in relation to genomic basis of local adaptation, emphasize the lack of large effect loci contributing to local adaptation in the coding regions and thus point out to the need for more attention towards multi-locus analysis of polygenic adaptation.

Homologous recombination DNA-repair deficiency (HRD) is a common driver of genomic instability and confers a therapeutic vulnerability in cancer. The accurate detection of somatic allelic imbalances (AIs) has been limited by methods focused on BRCA1/2 mutations and using mixtures of cancer types. Using pan-cancer data, we revealed distinct patterns of AIs in high-grade serous ovarian cancer (HGSC). We used machine learning and statistics to generate improved criteria to identify HRD in HGSC (ovaHRDscar). ovaHRDscar significantly predicted clinical outcomes in three independent patient cohorts with higher precision than previous methods. Characterization of 98 spatiotemporally distinct metastatic samples revealed low intra-patient variation and indicated the primary tumor as the preferred site for clinical sampling in HGSC. Further, our approach improved the prediction of clinical outcomes in triple-negative breast cancer (tnbcHRDscar), validated in two independent patient cohorts. In conclusion, our tumor-specific, systematic approach has the potential to improve patient selection for HR-targeted therapies.