@genebe/pangolin

Pangolin masked splice gain/loss scores

This annotation contains compact Pangolin scores for all precomputed GRCh38 SNVs in protein-coding genes. Pangolin predicts changes in splice-site usage; the source scores were calculated with the default 50-nucleotide window and masked using GENCODE splice-site annotations.

For each genomic SNV, the builder selects the source record with the largest raw max_score = max(gain_score, loss_score). Scores smaller than 0.1 are omitted with their corresponding relative position to reduce storage: gain_score and gain_pos are both NULL below the gain threshold, and the same rule applies to loss. max_score is always retained and is calculated before this thresholding.

Source and citation

Wagner N, Neverov A. Pangolin precomputed scores. Zenodo, version 1 (2025). https://doi.org/10.5281/zenodo.15649338

Zeng T, Li YI. Predicting RNA splicing from DNA sequence using Pangolin. Genome Biology. 2022;23:103. https://doi.org/10.1186/s13059-022-02664-4