The source code for this example can be found on GitHub

PolA-SODA was developed as a debugging tool for PolyA, which is a tool for adjudicating between competing annotations of biological sequences. The development of PolyA itself is specifically motivated by the adjudication of TE annotation. As such, this PolyA-SODA example application is set up to compare TE annotations as adjudicated by PolyA to existing TE annotations on the UCSC Genome Browser RepeatMasker Viz track.

Under the hood, PolyA uses a collection of overlapping sequence alignments between distinct features and a genome to compute position specific confidence scores for each feature. Simply put, the confidence scores are a measure of how confident we are that a position on the genome should be annotated with a given feature.

The input to PolyA-SODA is a JSON string that is optionally output by PolyA. You can download an example JSON here, or press the example button to render one in the application. br Once a JSON has been submitted, two soda charts displaying TE annotations for the same region in the human genome (hg38) will render. The top chart displays the TE annotations from the UCSC Genome Browser, and the bottom chart displays PolyA annotations. By clicking and dragging on the bottom chart, you can make a brush selection of a region, which render a heatmap of relevant annotation confidence scores and the sequence alignments they are calculated from.