Analysis Scripts for DigiPico Sequencing Data
MutLX tool for accurate identification of true islet specific variants from DigiPico data.
You can either download or clone this repository.
This project has been tested on Python 3. The MutLX/requirements.txt file contains all Python libraries that you need. They can be installed by running the following command in the project's folder:
pip install -r requirements.txt
The input should be a csv file in the following format without header:
Mutation, Type, [41 MutLX features extracted from DigiPico data]
Type column must be of any of the following categories:
- SNP-Unq: For UTD variants
- SNP-Hm: For homozygous germline variants
- SNP-Ht-H: For high-confidence heterezygous germline variants (Important for tumours with complex genomes)
- SNP-Ht: For other heterozygous germline variants
- SNP-Somatic: For known somatic mutations
Example files for analysis with DigiPico can be downloaded from below links:
- Sample file for DigiPico run without expected islet specific variants
- Sample file for DigiPico run with expected islet specific variants
You can run the following command in the project's folder:
python mutLX.py --input test1.csv --out_path test1_Results --sample_name DigiPico_test1
- --input: Path to input csv file
- --out_path: Output directory path (default = run directory)
- --sample_name: Sample name to be used as prefix for output files (default = "DigiPico")
- --batch_size: Training batch size (default = 8)
- --epochs: Number of epochs in training (default = 10)
- --subset_num: Number of training subsets to be used for training (default = 25)
- --drop_it: Number of iterations for dropout analysis (default = 100)
- --pscore_cf: Probability score cut-off value (default = 0.2)
- --auc_cf: Cut-off value for area under the ROC curve to identify samples with true UTDs (default = 0.9)
- --tpr_cf: The required true positive rate based on germline SNPs for the recovery of true UTDs (default = 0.95)
mutLX.py will generate a final sample_name_scores.csv file in the out_path directory with the below header as described in our manuscript:
Mutation, Type, Probability_Score, Uncertainty_Score, Result
It will also generate several plots to represent the data.
This analysis refers to the following publication: Eli M Carrami, Sahand Sharifzadeh, Nina C Wietek, Mara Artibani, Salma El-Sahhar, Tatjana Sauka-Spengler, Christopher Yau, Volker Tresp, Ahmed A Ahmed. A highly accurate platform for clone-specific mutation discovery enables the study of active mutational processes. Available at 10.7554/eLife.55207, eLife, 2020."