Sparse reduced?rank regression for exploratory visualisation of paired multivariate data

In genomics, transcriptomics, and related biological fields (collectively known as omics), combinations of experimental techniques can yield multiple sets features for the same set replicates. One example is Patch-seq, a method combining single-cell RNA sequencing with electrophysiological recordings from cells. Here we present framework based on sparse reduced-rank regression (RRR) obtaining an interpretable visualisation relationship between transcriptomic data. We use elastic net regularisation that yields solutions allows efficient computational implementation. Using several Patch-seq datasets, show RRR outperforms both full-rank non-sparse RRR, well previous approaches, in terms predictive performance. introduce bibiplot order to display dominant factors determining properties neurons. believe provide valuable tool exploration paired multivariate datasets.