{"id":27,"date":"2025-01-22T19:59:51","date_gmt":"2025-01-22T19:59:51","guid":{"rendered":"https:\/\/mrivas.su.domains\/gbe\/?p=27"},"modified":"2025-03-11T19:27:57","modified_gmt":"2025-03-11T19:27:57","slug":"efficient-regression-for-population-scale-genome-sequencing-studies","status":"publish","type":"post","link":"https:\/\/mrivas.su.domains\/gbe\/uncategorized\/efficient-regression-for-population-scale-genome-sequencing-studies\/","title":{"rendered":"Efficient regression for population scale genome sequencing studies"},"content":{"rendered":"\n<p>As we move from <em>Common Variant Association Studies (CVAS)<\/em> to <em>Rare Variant Association Studies (RVAS)<\/em> it has become increasingly obvious that the majority of our computational workload will be dedicated to analyzing rare variants. <\/p>\n\n\n\n<p>For simple illustration here is a side by side barplot showing the absolute number of common variants in a<strong> whole genome sequencing study <\/strong>of approximately 200,000 individuals (<a href=\"https:\/\/databrowser.researchallofus.org\/\">AllofUs dataset<\/a>).<\/p>\n\n\n\n<figure class=\"wp-block-image aligncenter size-full is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"666\" height=\"812\" src=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-4.png\" alt=\"\" class=\"wp-image-28\" style=\"width:195px;height:auto\" srcset=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-4.png 666w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-4-246x300.png 246w\" sizes=\"auto, (max-width: 666px) 100vw, 666px\" \/><\/figure>\n\n\n\n<figure class=\"wp-block-image aligncenter size-large is-resized\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"664\" src=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6-1024x664.png\" alt=\"\" class=\"wp-image-30\" style=\"width:460px;height:auto\" srcset=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6-1024x664.png 1024w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6-300x195.png 300w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6-768x498.png 768w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6-1536x997.png 1536w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-6.png 1697w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>If we were to visualize it in a barplot the difference is quite dramatic.<\/p>\n\n\n\n<p>I initially analyzed the body mass index phenotype in the AllofUs cohort using <a href=\"https:\/\/www.cog-genomics.org\/plink\/2.0\/\">PLINK 2.0<\/a>.<\/p>\n\n\n\n<p>PLINK 2.0 has computationally efficient algorithms. However, it was quite clear that the original algorithms required quite a lot of computational resources to run it across hundreds of thousands of individuals. <\/p>\n\n\n\n<p>To compare we conducted univariate regression across the exome for the body mass index (BMI) phenotype in the AllofUS cohort, which required 11.6 hours using 50 threads on a single machine.<\/p>\n\n\n\n<p>We realized that we could improve the efficiency of the study by taking advantage of the property that most of the variants that were being analyzed were <em><strong>rare<\/strong>. <\/em><\/p>\n\n\n\n<p>To compute the estimates of the regression coefficients we only needed access to the data from the rare variant carriers after <em>residualizing the covariates (this computation only needs to be done once)<\/em>.<\/p>\n\n\n\n<p>We were able to<strong> reduce<\/strong> computation time down from <strong>11.6 hours <\/strong>to <strong>1.6 minutes <\/strong>using 50 threads in a single machine.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"683\" src=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-1024x683.png\" alt=\"\" class=\"wp-image-31\" srcset=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-1024x683.png 1024w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-300x200.png 300w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-768x512.png 768w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-1536x1024.png 1536w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-7-2048x1365.png 2048w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<p>We also found that we were able to maintain power and control type 1 error rates.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"561\" height=\"432\" src=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-8.png\" alt=\"\" class=\"wp-image-32\" srcset=\"https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-8.png 561w, https:\/\/mrivas.su.domains\/gbe\/wp-content\/uploads\/2025\/01\/image-8-300x231.png 300w\" sizes=\"auto, (max-width: 561px) 100vw, 561px\" \/><\/figure>\n\n\n\n<p>Paper is found <a href=\"https:\/\/www.biorxiv.org\/content\/10.1101\/2024.04.11.589062v2\">here<\/a> on <a href=\"https:\/\/www.biorxiv.org\/content\/10.1101\/2024.04.11.589062v2\">biorxiv<\/a> and it is <em>In Press <\/em>at <em><a href=\"https:\/\/academic.oup.com\/bioinformatics\/advance-article\/doi\/10.1093\/bioinformatics\/btaf067\/8008994\">Bioinformatics<\/a><\/em>.<\/p>\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>As we move from Common Variant Association Studies (CVAS) to Rare Variant Association Studies (RVAS) it has become increasingly obvious that the majority of our computational workload will be dedicated to analyzing rare variants. For simple illustration here is a side by side barplot showing the absolute number of common variants in a whole genome [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-27","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/posts\/27","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/comments?post=27"}],"version-history":[{"count":5,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/posts\/27\/revisions"}],"predecessor-version":[{"id":227,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/posts\/27\/revisions\/227"}],"wp:attachment":[{"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/media?parent=27"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/categories?post=27"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/mrivas.su.domains\/gbe\/wp-json\/wp\/v2\/tags?post=27"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}