A preprint titled “Natural coordinate descent algorithm for L1-penalised regression in generalised linear models” is available on the arXiv. L1-penalised regression refers to the use of a penalty term that sums the absolute effect sizes of all parameters in a high-dimensional regression model to select the most important variables in the model. This paper uses a simple result from convex analysis to show that a solution mechanism (“soft-thresholding”) previously known for least-squares regression is in fact generic and also holds for regression with for instance binary or count-based data.
A preprint of our paper “Graph-based data integration predicts long-range regulatory interactions across the human genome” is available on arXiv and bioRxiv. In this paper we use graph-based methods to combine multiple datasets of open chromatin and gene expression information in matching cell types from the ENCODE project to predict long-range interactions between regulatory elements in the human genome and their target genes. Predicted interactions between can be browsed here.
A preprint of our paper kruX: Matrix-based non-parametric eQTL discovery is available from the arXiv. This paper describes a software tool called kruX for performing millions of non-parametric ANOVA (Kruskal-Wallis) tests at once using matrix-multiplication methods. kruX is about 3 orders of magnitude faster than performing these tests one-by-one, which makes a difference if you want to do billions of them!