Warning: This project is in beta stage. It is under active development and may be unstable.
Convert human T-cell receptor (TCR) annotations between 10X, Adaptive, and IMGT formats.
The most popular TCR sequencing platforms (Adaptive and 10X) and the widely used IMGT reference all use different naming conventions for their TCR gene annotations. These are all the same gene:
- 10X:
TRAV1-2 - Adaptive:
TCRAV01-02*01 - IMGT:
TRAV1-2*01
tcrconvert easily converts TCR annotations between formats.
Check out the full documentation here: tcrconvert.readthedocs.io
tcrconvert requires pandas.
You can install using pip. Clone this repo, then from the top-level folder run:
pip install .
The lookup tables for translating gene names come pre-built from IMGT fasta files located under tcrconvert/data/
Load your 10X, Adaptive, or IMGT-formatted TCR data into a pandas dataframe. It may have come from files such as:
- 10X:
filtered_contig_annotations.csv - Adaptive:
SAMPLE_TCRB.tsv - IMGT: A custom CSV file
Then, convert to your desired format:
import pandas as pd
import tcrconvert
df = pd.read_csv("filtered_contig_annotations.csv")
converted = tcrconvert.convert_gene(df, frm='tenx', to='adaptive') # From 10X to Adaptive
Availble formats to convert to or from are: tenx, adaptive, adaptivev2, or imgt
To use Adaptive Immune Receptor Repertoire (AIRR) formatted files, simply specify the column names:
import pandas as pd
import tcrconvert
df = pd.read_csv("airr_rearrangement.tsv", sep='\t')
converted = tcrconvert.convert_gene(df, frm='tenx', to='adaptive',
frm_cols=['v_call', 'd_call', 'j_call', 'c_call'])
I welcome feedback! If you would like to resolve an issue or add improvements please submit a pull request.
If you run into problems or need help running TCRconvert please file an issue on GitHub.
For other questions please contact Emma Bishop: emmab5 at uw dot edu