VARiD

VarID is a computational tool to analyze cell state transitions in single-cell transcriptomes, focusing on the challenge of quantifying weak expression of lineage-determining factors. It identifies locally homogenous neighborhoods within cell state space to effectively measure local gene expression variability. By delineating these neighborhoods, VarID can uncover the pseudo-temporal dynamics of gene expression variability during cell differentiation. This approach is crucial for understanding the nuanced changes in gene expression that guide cell fate decisions.

Additionally, VarID is critical in genomic variation studies by integrating data from high-throughput sequencing (HTS) technologies, including traditional nucleotide sequencing and the diabase-coded (color space) sequences generated by platforms like the Applied Biosystem's solid. It employs a probabilistic model based on a hidden Markov model. It utilizes the forward-backward algorithm to identify genetic variants, such as heterozygous, homozygous, tri-allelic SNPs, and micro-indels from both letter- and color-space reads simultaneously.