SAIGE-GPU: accelerating genome- and phenome-wide association studies using GPUs.

Alex Rodriguez; Youngdae Kim; Tarak Nath Nandi; Karl Keat; Rachit Kumar; Mitchell Conery; Rohan Bhukar; Molei Liu; John Hessington; Ketan Maheshwari; VA Million Veteran Program; Edmon Begoli; Georgia Tourassi; Sumitra Muralidhar; Pradeep Natarajan; Benjamin F Voight; Kelly Cho; John Michael Gaziano; Scott M Damrauer; Katherine P Liao; Wei Zhou; Jennifer E Huffman; Anurag Verma; Ravi K Madduri

doi:10.1093/bioinformatics/btag032

Home/Archive/PubMed

PubMed · 41572430

SAIGE-GPU: accelerating genome- and phenome-wide association studies using GPUs.

View original source ↗Export metadata ↓

Abstract

MOTIVATION: Genome-wide association studies (GWAS) at biobank scale are computationally intensive, especially for admixed populations requiring robust statistical models. SAIGE is a widely used method for generalized linear mixed-model GWAS but is limited by its CPU-based implementation, making phenome-wide association studies impractical for many research groups. RESULTS: We developed SAIGE-GPU, a GPU-accelerated version of SAIGE that replaces CPU-intensive matrix operations with GPU-optimized kernels. The core innovation is distributing genetic relationship matrix calculations across GPUs and communication layers. Applied to 2068 phenotypes from 635 969 participants in the Million Veteran Program, including diverse and admixed populations, SAIGE-GPU achieved a 5-fold speedup in mixed model fitting on supercomputing infrastructure and cloud platforms. We further optimized the variant association testing step through multi-core and multi-trait parallelization. Deployed on Google Cloud Platform and Azure, the method provided substantial cost and time savings. AVAILABILITY AND IMPLEMENTATION: Source code and binaries are available for download at https://github.com/saigegit/SAIGE/tree/SAIGE-GPU-1.3.3. A code snapshot is archived at Zenodo for reproducibility (DOI: [10.5281/zenodo.17642591]). SAIGE-GPU is available in a containerized format for use across HPC and cloud environments and is implemented in R/C++ and runs on Linux systems.

Explore related subjects

Genome-Wide Association Study Software Phenotype Humans Phenomics Algorithms Journal Article Research Support, Non-U.S. Gov't

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Alex Rodriguez, Youngdae Kim, Tarak Nath Nandi, Karl Keat, Rachit Kumar, Mitchell Conery, Rohan Bhukar, Molei Liu, John Hessington, Ketan Maheshwari, VA Million Veteran Program, Edmon Begoli, Georgia Tourassi, Sumitra Muralidhar, Pradeep Natarajan, Benjamin F Voight, Kelly Cho, John Michael Gaziano, Scott M Damrauer, Katherine P Liao, Wei Zhou, Jennifer E Huffman, Anurag Verma, Ravi K Madduri. 2026-02-28. SAIGE-GPU: accelerating genome- and phenome-wide association studies using GPUs.. https://doi.org/10.1093/bioinformatics%2Fbtag032

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2026

Pinpointing genomic regions conferring herbicide tolerance in cassava via genome-wide association mapping.

Cassava (Manihot esculenta Crantz) is a tropical crop of major socioeconomic importance, whose productivity can be limited by sensitivity to herbicides used for weed management. This study aimed to perform a genome-wide association study (GWAS) in 194 cassava genotypes to identify genomic regions associated with tolerance to the herbicides mesotrione, S-metolachlor, and chloransulam-methyl. The evaluations performed at 3, 6, 9, 15, and 30 days after application (DAA) were used to characterize the temporal progression of phytotoxicity. Based on this analysis, the phenotype obtained at 9 days after application (PhytoX9DAA) was selected for genome-wide association analyses because it represented the period of greatest symptom expression and the highest discrimination among genotypes. GWAS analyses were performed using de-regressed BLUPs and the MLM, MLMM, and BLINK models, incorporating kinship (K) and population structure (Q) matrices. Significant markers were detected across multiple chromosomes, and the corresponding genomic windows contained candidate genes with functional annotations related to herbicide response. The predominant functional categories included membrane transport, channel activity, signal peptide processing, protein phosphorylation, cellular signaling, and metabolic regulation. Key candidate genes included Manes.02G151900 and Manes.02G152700 (chromosome 2), associated with transmembrane transport and signal peptide processing; Manes.09G060900 (chromosome 9), associated with protein kinase activity, ATP binding, and protein phosphorylation; and Manes.15G083800 and Manes.15G084000 (chromosome 15), associated with S-adenosylmethionine-dependent methyltransferase activity, membrane-related functions, and protein phosphorylation. These genes participate in biochemical pathways involved in cellular signaling, membrane transport, and metabolic regulation that may contribute to herbicide tolerance. Overall, the results demonstrate that herbicide tolerance in cassava is a quantitative and polygenic trait governed by numerous small-effect loci. The integration of cellular signaling, metabolic regulation, and membrane transport supports the physiological resilience of the species under chemical exposure, providing valuable insights for breeding strategies and marker-assisted selection.

Genome-Wide Association Study↗

PubMed2026

Dissecting genetic architecture of growth and yield traits in horsegram using GWAS.

Horsegram (Macrotyloma uniflorum), a member of the Fabaceae family, is a nutritious and low-cost legume used for both grain and fodder. This study employed a genome-wide association approach to identify loci linked to key agronomic traits in horsegram. Plant height, seed size, and shoot fresh weight were evaluated in a panel of 96 diverse genotypes. GBS was performed using the Illumina HiSeq platform, yielding 20,241 high-quality SNPs after filtering at a 5% minor allele frequency. Population structure analysis classified genotypes into three admixed subgroups. Phenotyping was conducted over three consecutive years at two locations in Himachal Pradesh (Palampur and Bajaura) using a randomized block design with two replications. GWAS analyses using GLM, MLM, FarmCPU, and BLINK models identified eight markers for plant height, three for seed size, and five for shoot fresh weight across different chromosomes. These markers provide valuable tools for accelerating trait improvement in future horsegram breeding programs.

Genome-Wide Association Study↗

PubMed2026

A module-based approach for post-omics, post-GWAS network-based gene classification.

MOTIVATION: Complex traits and diseases are highly polygenic and understanding the full set of genes involved is a central challenge in biomedicine. However, due to sample size limitations and noise (technical and biological), experimental approaches for disease-gene discovery such as transcriptomics and GWAS result in long, noisy, heterogeneous gene lists, which may be trimmed to a subset of likely relevant genes while leaving several false negatives. Computational gene classification approaches, especially those using genome-scale molecular interaction networks, are promising avenues for complementing such experimental findings by analytically expanding observed gene lists based on the functional relatedness between genes. We previously introduced the network-based gene classification approach, GenePlexus, which was rigorously benchmarked to show state-of-the-art performance, especially for predicting novel genes associated with biological processes and fine-grained phenotypes. Network-based gene classification performance,however, declines for diseases, especially when the inputs are omics and GWAS-based long gene lists. RESULTS: Here, we show that these disease gene lists span multiple biological processes spread across the molecular network, and we propose ModGenePlexus, a new network-based gene classification method that takes a two-stage approach. First, clustering and semi-supervised learning decomposes the input gene list into coherent, denoised network gene modules. Then, ModGenePlexus trains supervised (GenePlexus) classifiers for each module and aggregates predictions to return genome-wide rankings. We benchmarked ModGenePlexus across simulated data, transcriptomic signatures, and GWAS datasets (together spanning hundreds of diseases), showing improved recovery of known disease genes compared to GenePlexus. Beyond improved classification, the results of enrichment analysis of ModGenePlexus outputs are much more interpretable by virtue of revealing nuanced biological processes. Together, these results establish ModGenePlexus as a scalable, interpretable tool for gene classification of GWAS- and omics-derived gene lists across diverse biological contexts. AVAILABILITY AND IMPLEMENTATION: ModGenePlexus is freely available on GitHub at https://github.com/krishnanlab/ModGenePlexus, and the full source code and results supporting this study are available on Zenodo at https://zenodo.org/records/19857910.

Genome-Wide Association Study↗