criTRia

criTRia (pronounce criteria) is a formalized scoring framework designed to accurately evaluate TR-disease relationships. criTRia builds on current best-practice scoring procedures developed by the Clinical Genome Resource (ClinGen) while (re)curating based on locus-level (rather than gene-level) classification, introducing TR-specific evidence categories and reweighted scoring. In this repository you will find all data and code used to evaluate success and create the figures for the manuscript and SOP, up to date as of the latest commit.

Download and format Google Sheet (Python)

Use download_sheet_to_criteria_dataset.py to download a Google Sheet tab and standardize it to the same format as criTRia_Dataset.csv:

• Output columns: Gene,Group,categorical_score
• Refuted and Disputed scores are converted to Contradictory

Run from the repository root:

python3 download_sheet_to_criteria_dataset.py

Merge data and generate figures

Rscript criTRia_Figure_Script.R

Figure source files

This repository includes links to editable source files for figures used in the criTRia manuscript.

Figure #	Description	Editable Source Link
Figure 1	criTRia scoring framework overview	https://docs.google.com/spreadsheets/d/1VEuZqvwtQWzVSBc7Aj4FUPQABwyNTCcb8XDNCZ-xoKM/edit?usp=sharing
Figure 2	Comparison of criTRia vs Gene Curation Coalition	https://docs.google.com/spreadsheets/d/1DonSiPVjeQLsB8HoFzn50jFCnWvMPxVZ124SzS0s7r0/edit?gid=0#gid=0
Figure 3	TR-disease association results	https://github.com/dashnowlab/criTRia/blob/main/criTRia_Figure_Script.R

Repository contributors:

• Principal Investigator: Harriet Dashnow
• First Author: Macayla Ann Weiner