TACaPe: Transformed-based Anti-Cancer Peptide Classification and Generation

TACaPe (Transformed-based Anti-Cancer Peptide Classification and Generation) is a commandline tool to train transformer-based models for anticancer peptide classification and generation. I was built on top of Tensorflow and uses an auto-regressive algorithm for peptide design, which results can be filtered using an optional classification model.

Setup

Installing from PyPI using pip

$ pip install tacape

Installing from GitHub

$ git clone https://github.com/omixlab/anticancer-peptide
$ cd anticancer-peptide

Using pip

$ pip install -r requirements.txt -e .

Using conda

$ conda env create
$ conda activate anticancer-peptide

Usage

tacape-train-classifier

usage: TACaPe: Model Training [-h] --positive-train POSITIVE_TRAIN --negative-train NEGATIVE_TRAIN --positive-test POSITIVE_TEST
                              --negative-test NEGATIVE_TEST [--format {text,fasta}] --output OUTPUT [--epochs EPOCHS]

optional arguments:
  -h, --help            show this help message and exit
  --positive-train POSITIVE_TRAIN
                        Input file containing positive peptides for training
  --negative-train NEGATIVE_TRAIN
                        Input file containing negative peptides for training
  --positive-test POSITIVE_TEST
                        Input file containing positive peptides for testing
  --negative-test NEGATIVE_TEST
                        Input file containing negative peptides for testing
  --format {text,fasta}
                        [optional] Input file format (default: text)
  --output OUTPUT       Path prefix of the output files
  --epochs EPOCHS       [optional] Number of epochs to be used during training (default: 30)

tacape-predict

usage: TACaPe: Predict [-h] --input INPUT [--format {text,fasta}] --classifier-prefix CLASSIFIER_PREFIX --output OUTPUT

optional arguments:
  -h, --help            show this help message and exit
  --input INPUT         Input file
  --format {text,fasta}
                        [optional] Input file format (default: text)
  --classifier-prefix CLASSIFIER_PREFIX
                        [optional] Path to the file prefix of the trained classification model
  --output OUTPUT       Path to the output CSV file