Abstract | ||
---|---|---|
The availability of large-scale genomic and transcriptomic data on populations makes it necessary to perform computationally intensive expression quantitative trait locus (eQTL) analysis. Modeling in a sparse learning framework, LASSO based tools are powerful for eQTL analysis. However, classical LASSO becomes limited for big genomic data. We thus propose two novel methods, namely sequential LASSO and parallel LASSO, to conduct eQTL analysis for datasets of ultra-high dimension. We theoretically prove the consistency of our methods under mild conditions and perform extensive simulations on synthetic data to validate our methods. We also apply our methods to a real human genomics database demonstrate the application of our method. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1145/2808719.2811449 | BCB |
Field | DocType | Citations |
Computer science,Lasso (statistics),Genomics,Synthetic data,Artificial intelligence,Expression quantitative trait loci,Bioinformatics,Machine learning,Sparse learning | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Anhong He | 1 | 0 | 0.34 |
Benika Hall | 2 | 2 | 1.48 |
Jia Wen | 3 | 0 | 1.35 |
Yingbin Liang | 4 | 1646 | 147.64 |
Xinghua Shi | 5 | 209 | 19.00 |