2026-10-14T06:45:58Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33976971 2026-09-23T17:50:55Z category_13 category_24 category_734 category_811 category_106 category_132 portal_5 item_type_3 month_year_09_2026

<p>Number of non-zero coefficients for each phenotype measured across all Lasso-based algorithms. Compared to Lasso, uniLasso selects 38% fewer SNPs for height and 43% fewer for BMI, while only reducing <i>R</i>² by 0.006 and 0.016, respectively.</p> Joshua Richland (25099428) Tuomo Kiiskinen (8704011) William Wang (5811092) Wenhui Sophia Lu (25099431) Balasubramanian Narasimhan (6444833) Trevor Hastie (195) Manuel Rivas (268944) Robert Tibshirani (196) Genetics Evolutionary Biology Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Science Policy Infectious Diseases stabilize feature selection regression toward variants level target data informing penalty weights guided sparse regression based repository comprising external statistics guide external scores ). recently introduced univariate leverages univariate coefficients prs estimation methods signals guide unilasso ). yielding sparser xlink "> united kingdom standard lasso sign constraints results demonstrate remains competitive prior evidence interpretable models empirical advantages <p>Number of non-zero coefficients for each phenotype measured across all Lasso-based algorithms. Compared to Lasso, uniLasso selects 38% fewer SNPs for height and 43% fewer for BMI, while only reducing <i>R</i>² by 0.006 and 0.016, respectively.</p> 2026-09-23T17:43:15Z Dataset Dataset 10.1371/journal.pgen.1012314.t001 https://figshare.com/articles/dataset/_p_Number_of_non-zero_coefficients_for_each_phenotype_measured_across_all_Lasso-based_algorithms_Compared_to_Lasso_uniLasso_selects_38_fewer_SNPs_for_height_and_43_fewer_for_BMI_while_only_reducing_i_R_i_sup_2_sup_by_0_006_and_0_016_respect/33976971 CC BY 4.0