SArKS:de novodiscovery of gene expression regulatory motifs and domains by suffix array kernel smoothing

Abstract

Experiments designed to assess differential gene expression represent a rich resource for discovering how DNA regulatory sequences influence transcription. Results derived from such experiments are usually quantified as continuous scores, such as fold changes, test statistics and p-values. We present a de novo motif discovery algorithm, SArKS, which uses a nonparametric kernel smoothing approach to identify promoter motifs correlated with elevated differential expression scores. SArKS has the capability to smooth over both motif sequence similarity and, in a second pass, over spatial proximity of multiple motifs to identify longer regions enriched in correlative motifs. We applied SArKS to simulated data, illustrating how SArKS can be used to find motifs embedded in random background sequences, and to two published RNA-seq expression data sets, one probing S. cerevisiae transcriptional response to anti-fungal agents and the other comparing gene expression profiles among cortical neuron subtypes in M. musculus. For both RNA-seq sets we successfully identified motifs whose kernel-smoothed scores were significantly elevated compared to the permutation-estimated background distributions. We found strong similarities between these identified motifs and known, biologically meaningful sequence elements which may help to provide additional context for the results previously published regarding these data sets. Finally, because eukaryotic transcription regulation is highly combinatorial, we also outline how SArKS methods might be extended to discover synergistic motifs.

Functional Access to Neuron Subclasses in Rodent and Primate Forebrain

“…We then used our recently developed algorithm, SArKS (Wylie et al, 2018), and mined RNA sequencing (RNA-seq) data (Mo et al, 2015) for sequence motifs associated with cell type-specific expression in PV + neurons. Among the genes highlighted by SArKS was PaqR4, a member of the progestin receptor family (Tang et al, 2005).…”

“…In another example, we developed an algorithm that searched de novo for DNA motifs shared by putative promoters of genes expressed in mouse PV + neurons but not VIP + or excitatory neurons (Wylie et al, 2018). Many of the uncovered motifs were not present in repositories of transcription factor binding sites (Khan et al, 2018).…”

“…We then set out to develop a general and rational approach for promoter candidate selection that aimed to minimize the hit-or-miss aspect of existing strategies. Our goal was to apply our recently developed Suffix Array Kernel Smoothing algorithm, (SArKS; Wylie et al, 2018), that was designed to mine the growing body of RNA-seq data for sequence motifs associated with cell type-specific gene expression.…”

“…The PV gene itself fulfilled most of the enumerated criteria, but its chromatin was differentially accessible (Mo et al, 2015); the PV promoter was consequently eliminated from contention. The upstream regions (~3 kb) of the remaining 6,326 genes were examined using SArKS (Wylie et al, 2018) to find motifs (k-mers) whose occurrence in a set of promoter sequences correlated with an input metric of differential expression: a t-statistic comparing the TPM-normalized RNA transcript abundance in PV + neurons versus PV − neurons. SArKS first identified motifs by employing smoothing over subsequences by sequence similarity and then identified multi-motif domains (MMDs) by additionally smoothing over spatial proximity, using a permutation testing approach to establish statistical significance.…”

“…(A) SArKS-facilitated selection ofthe PaqR4 gene (Wylie et al, 2018). Cell class-specific transcriptome data (Mo et al, 2015) were hierarchically clustered on the basis of differential expression in PV + , VIP + , and excitatory (EXC) neurons (as shown in dendrogram).…”