Quantitative prediction model of human RNA transcript coding ability

Recently, the Bioinformatics Center (CBI) of Peking University School of Life Sciences (CBI), Peking University Biomedical Frontier Innovation Center (BIOPIC) and Beijing Future Gene Diagnosis Advanced Innovation Center (ICG) Gao Ge Researcher Group, based on heterogeneous multi-omics data, A cross-cell quantitative model of human RNA transcript coding ability (coding ability) was established

The transcription of genes plays a role in linking the previous and the next in the central principles of biology.

The research team collected human Ribo-seq/RNA-seq paired data from 22 different cell types published in public databases and performed systematic mining and analysis, and strictly determined the translation status of 101,170 transcripts in the data

On this basis, the author applied a data-driven feature selection algorithm, integrated the use of sequence intrinsic (cis-) and cellular environment (contextual, trans-) features to establish a cross-cell quantitative model of human RNA transcript coding ability, RiboCalc, The high-precision prediction (r=0.

Figure 1: RiboCalc workflow and analysis results: A.

As the first quantitative prediction model of transcript coding ability developed for higher mammals, RiboCalc has realized the quantitative modeling and prediction of transcript coding ability across tissues and cell types in humans

Postdoctoral fellow at the School of Life Sciences, Peking University, Yujian Kang is the first author of the paper, and Gao Ge is the corresponding author.

references:

1.