This repository includes code for the analyses described in:

"Mutational scanning of CRX classifies clinical variants and reveals biochemical properties of the transcriptional effector domain" (10.1101/gr.279415.124)

James L. Shepherdson¹﹐²
David M. Granas¹﹐²
Jie Li¹﹐²
Zara Shariff¹﹐²
Stephen P. Plassmeyer³﹐⁴
Alex S. Holehouse³﹐⁴
Michael A. White¹﹐²
Barak A. Cohen¹﹐²^

¹Department of Genetics,
²Edison Family Center for Genome Sciences & Systems Biology,
³Department of Biochemistry and Molecular Biophysics,
⁴Center for Molecular Condensates,
Washington University in St. Louis School of Medicine, St. Louis, MO 63110, USA

^ Correspondence: Barak A. Cohen [email protected]

Description

• DMS_analysis: includes code and scripts for analysis of barcode abundance data from sorted fractions to compute variant activity scores and generate figures
  • call_variants.py generates a barcode-to-variant map from PacBio long read sequencing data to associate sequence barcodes with coding variants
• tools: includes software tools developed for this manuscript that are used in other scripts:
  • bcbuddy is used to extract barcodes from sequencing reads
  • dms_tools includes a library for parsing the output of minimap2 for long read data analysis; used by call_variants.py
  • fphd implements a parallelized Hamming distance calculation for barcode error correction

Notes

crx_genomic_positions.tsv is provided to translate protein-level variant coordinates into cDNA and gDNA coordinates. To produce a VCF-compatible file, awk or a similar tool can be used. For example:

awk 'BEGIN {FS="\t"; OFS="\t"} {print 19, $5, ".", $6, $7, ".", "PASS", "var="$1$2$3}' crx_genomic_positions.tsv

Note that you may wish to first filter the crx_genomic_positions table to particular variants of interest, or join it with the DMS or computational predictor data to include scores in the VCF.