Decoy-Free FDR Estimation for Mass-Spectrometry

A package to estimate FDR in mass-spectrometry searching results using decoy-free approach

Install

pip install decoyfree-msfdr

Usage

Regular MS/MS search FDR estimation

decoyfree-msfdr [-h] [-f INPUT_FILE] [-d INPUT_DIR] [-t INPUT_TYPE] [-m EVAL_MODEL]
                [--score-field SCORE_FIELD] [--threads THREADS] [-c CONSTRAINTS] [-s MODEL_SAMPLES]
                [-r RANDOM_SIZE] [--tolerance TOLERANCE] [--show_plotting] [--out_dir OUT_DIR]

Cross-linked peptide MS/MS search FDR estimation

decoyfree-xlmsfdr [-h] [-f INPUT_FILE] [-d INPUT_DIR] [-t INPUT_TYPE] [-m EVAL_MODEL]
                  [--score-field SCORE_FIELD] [--threads THREADS] [-c CONSTRAINTS] [-s MODEL_SAMPLES]
                  [-r RANDOM_SIZE] [--tolerance TOLERANCE] [--show_plotting] [--out_dir OUT_DIR]

options:
  -h, --help            show this help message and exit
  -f INPUT_FILE, --input-file INPUT_FILE
  -d INPUT_DIR, --input-dir INPUT_DIR
  -t INPUT_TYPE, --input-type INPUT_TYPE
                        format of input files, supported formats are csv,tsv,idXML
  -m EVAL_MODEL, --eval-model EVAL_MODEL
                        existing model to be evaluated
  --score-field SCORE_FIELD
                        the field name holding PSM scores
  --threads THREADS     number of threads
  -c CONSTRAINTS, --constraints CONSTRAINTS
                        Choices of constraints to be used
  -s MODEL_SAMPLES, --model_samples MODEL_SAMPLES
                        Number of samples/top scores to be used in modeling
  -r RANDOM_SIZE, --random_size RANDOM_SIZE
                        Number of random starts per skewness setting
  --tolerance TOLERANCE
                        Threshold of the change of the point-wise log-likelihood for the EM algorithm to determine the
                        convergence
  --show_plotting       Show plotting while fitting the model
  --out_dir OUT_DIR     The place to save results

Options

Examples

Regular MS/MS search with MSGF+ engine

Suppose the MS/MS search result with MSGF+ software is saved in .tsv format, data/sample.tsv. You can run the FDR estimation algorithm with the following command,

decoyfree-msfdr -f data/sample.tsv -t tsv --out_dir results/sample --threads 10

Multiple search results

If you have multiple search results from MSGF+ saved in the 'data/sample/' directory, and you want to use them all together to build a single model, do the following,

decoyfree-msfdr -d data/sample/ -t tsv --out_dir results/sample --threads 10

Note: this will search the directory for all the files with .tsv extension. If you specified other formats, it will search for the files with the corresponding extension.

MS/MS search results with other engines

If you are using another search engine, please specify the following information,

XL-MS/MS search with OpenPepXLLF engine

Suppose the MS/MS search result with MSGF+ software is saved in .idXML format, data/sample.idXML. You can run the FDR estimation algorithm with the following command,

decoyfree-xlmsfdr -f data/sample.idXML -t idXML --out_dir results/sample --threads 10

Note: Currently, idXML is the only format we support. Please let us know if you need to use another format in issues. I'll add support to that.

XL-MS/MS search results with other engines

If you are using another search engine, please specify the following information,