Accurate genome-wide predictions of spatio-temporal gene expression during embryonic development

<p>This upload contains the expression prediction dataset discussed in the manuscript &quot;Accurate genome-wide predictions of spatio-temporal gene expression during embryonic development&quot; and used by&nbsp;the webserver&nbsp;https://find.princeton.edu.</p> <p>Abstract:</p> <p>Comprehensive information on the timing and location of gene expression is fundamental to our understanding of embryonic development and tissue formation.&nbsp; While high-throughput&nbsp;<em>in situ</em>&nbsp;hybridization projects provide invaluable information about developmental gene expression patterns for model organisms like&nbsp;<em>Drosophila</em>, the output of these experiments is primarily qualitative, and a high proportion of protein coding genes and most non-coding genes lack any annotation.&nbsp; Accurate data-centric predictions of spatio-temporal gene expression will therefore complement current <em>in situ</em>&nbsp;hybridization efforts.&nbsp; Here, we applied a machine learning approach by training models on all public gene expression and chromatin data, even from whole-organism experiments, to provide genome-wide, quantitative spatio-temporal predictions for all genes.&nbsp; We developed structured&nbsp;in silico&nbsp;nano-dissection, a computational approach that predicts gene expression in &gt;200 tissue-developmental stages. The algorithm integrates expression signals from a compendium of 6,378 genome-wide expression and chromatin profiling experiments in a cell lineage-aware fashion.&nbsp; We systematically evaluated our performance via cross-validation and experimentally confirmed 22 new predictions for four different embryonic tissues.&nbsp; The model also predicts complex, multi-tissue expression and developmental regulation with high accuracy.&nbsp; We further show the potential of applying these genome-wide predictions to extract tissue specificity signals from non-tissue-dissected experiments, and to prioritize tissues and stages for disease modeling.&nbsp; This resource, together with the exploratory tools are freely available at our webserver&nbsp;<a href="http://find.princeton.edu/">http://find.princeton.edu</a>, &nbsp;which provides a valuable tool for a range of applications, from predicting spatio-temporal expression patterns to recognizing tissue signatures from differential gene expression profiles.</p>