bulk-tumour-api: a programmatically accessible dataset of pre-processed bulk tumour sequencing data

<p><strong>This repository, including the API,&nbsp;are&nbsp;currently under development.</strong></p> <p><strong>bulk-tumour-api</strong>: A programmatically accessible dataset of pre-processed bulk tumour sequencing data. The python API can be found at&nbsp;https://github.com/tomouellette/bulk-tumour-api. All data stored in this repository have&nbsp;been collected from&nbsp;open access&nbsp;online sources. Original references and sources are provided in database.tsv (for empirical patient data) and synthetic.tsv (for simulated data).</p> <p><strong>A note on datasets: </strong></p> <ul> <li>All <em>empirical patient sequencing </em>samples&nbsp;have&nbsp;been processed into&nbsp;pseudo-VCF files&nbsp;which at minimum contain the following columns:&nbsp; sample identifier (sample), patient identifier (patient), chromosome (chr), position (pos), variant allele frequency (VAF), alternate read counts (t_alt_count), depth (DP), and total copy number (total_cn). However, if more data is required, unprocessed data including copy number segments or gene-level calls, clinical, and/or biopsy level information can be found in the /raw/.&nbsp;</li> <li>All <em>synthetic datasets </em>have also been processed in pseudo-VCF files. In some cases, all ground truth information (e.g. subclone frequency) is contained within the pseudo-VCF. In other cases, additional meta/ground-truth information are in separate files; any simulated sample with a column marked has_meta&nbsp;= True will have multiple files that will be downloaded together.</li> </ul>