pySAR

pySAR is a Python library for analysing Sequence Activity Relationships (SARs) of protein sequences. pySAR offers extensive and verbose functionalities that allow you to numerically encode a dataset of protein sequences using a large abundance of available methodologies and features. The software uses physiochemical and biochemical features from the Amino Acid Index (AAI) database as well as allowing for the calculation of a range of structural protein descriptors.
After finding the optimal technique and feature set at which to encode your dataset of sequences, pySAR can then be used to build a predictive regression model with the training data being that of the encoded sequences and training labels being the experimentally pre-calculated activity values for each protein sequence. The model can then be used to predict the activity/fitness value of a new unseen sequence.

status

Development Stage

To DO List:

• Add Github Workflow CI thing
• Add Category and Descriptor Group to pySAR results DF.
• Condense comments in functions, remove some whitespace lines
• Add help function
• Mention that PyBioMed package duplicated here as it is not available via pyPI and would mean that user would have to install the full pybiomed zip
• raise type errors instead of Value ?
• index errors?
• remove plot func from DSP
• do StanardScaler after every AAIndex encoding and before model building####
• Change importing globals : import globals / globals.OUTPUT_DIR
• Split up autocorrelation descriptors into their own functions
• Allow fasta file to be input to Descriptor class?
• github workflow with Twine that automatically published to pypi
• provide example script for running on GCP or AWS resources?
• don't return None after raising an exception??
• add descriptions to each methods in each class
• remove spacing in equals in keyword args in class/function defintiion
• setters and getters to Evaluate class? using @property
• add python version badge to readme
• add pypi badge to readme
• add introduction to readme
• add references to descriptor module
• integrate descriptor and AAIndex when using properties from AAIndex
• look into setup.cfg or setup.py
• add distance matrices json to dara ? : https://github.com/MartinThoma/propy3/blob/master/propy/QuasiSequenceOrder.py
• split up QuasiSequenceOrder descriptor into its consitent quasi-seq-order
• in readme show example usage for each module/class
• change AAI method names from get_feature etc to get_record...
• change get_feature_names to get_feature_desc
• add AAI category to each AAI record
• change all 'aa_index' to 'aaindex'
• Add assertion comments to each unit test, got X wanted Y..
• add test numbers in comments for each block of unit tests.
• Go through each parameters list and refer to its previous reference rather than repeating it.
• add cutoff index/value again just for testing
• print out default parameters if using them.
• remove verbose argument - dont need since tqdm prints progress bar
• add if name == "main" to encoding and pySAR class.
• split function defs to two lines?
• publish to conda?
• pypi logo
• license logo
• leave = False on 2nd loop

Installation

Install using pip:

pi3 install pySAR

Usage

Building predictive model from AAI and or protein descriptors:

e.g the below code will build a PlsRegression model using the AAI index CIDH920105 and the amino acid composition descriptor. The index is passed through a DSP pipeline and is transformed into its informational protein spectra using the power spectra, with a hamming window function applied to the output of the FFT. spectrum after a window function is applied.

#first-party imports
from globals import OUTPUT_DIR, OUTPUT_FOLDER, DATA_DIR
from aaindex import  AAIndex
from model import Model
from proDSP import ProDSP
from evaluate import Evaluate
import utils as utils
from plots import plot_reg
import descriptors as desc

pySAR = PySAR(dataset="dataset.txt",seq_col="sequence", activity="activity",algorithm = "PlsRegression", parameters={}, test_split=0.2)

results_df = pySAR.encode_aai_desc(indices="CIDH920105", descriptors="aa_composition", spectrum="power", window="hamming")

Encoding using all 566 AAIndex indices

#create instance of Encoding class, inherits from pySAR class
encoding = Encoding(dataset="dataset.txt", activity="activity_col",
  algorithm="RandomForest", parameters={"n_estimators":"200","max_depth":"50"})

aai_encoding = encoding.aai_encoding(spectrum='imaginary', window='blackman')

Encoding using list of 4 AAIndex indices, with no DSP functionalities

encoding = Encoding(dataset="dataset.txt", activity="activity_col",
  algorithm="PLSRegression", parameters={"":"","":"", })

aai_encoding = encoding.aai_encoding(use_dsp=False, aai_list=["PONP800102","RICJ880102","ROBB760107","KARS160113"])

Encoding using protein descriptors

encoding = Encoding(dataset="dataset.txt", activity="activity_col",
  algorithm="RandomForest", parameters={"":"","":"", }, descriptors_csv="descriptors.csv")

desc_encoding = encoding.desc_encoding(desc_combo = 2, verbose = True)
def descriptor_encoding(self, desc_list=None, desc_combo=1, verbose=True):

Encoding using AAI + protein descriptors

Generate all protein descriptors

  desc = Descriptor(protein_seqs = data, desc_dataset = "descriptors.csv",
      all_desc=True)

where protein_seqs is the dataset of protein sequences, desc_dataset is the name of the ouput csv used to store the calculated descriptors of the protein sequences and all_desc means that the class will get and calculate all descriptors.

Get record from AAIndex database

  desc = Descriptor(protein_seqs = data, desc_dataset = "descriptors.csv",
      all_desc=True)

Output Results

| Descriptor | Index | | R2 | RMSE | MSE
| ------------- | ------------- | | Content Cell | Content Cell | | Content Cell | Content Cell |

System Requirements

Python > 3.6 numpy >= 1.16.6 pandas >= 1.1.0 scikit-learn >= 0.24 scipy >= 1.4.1

Running Tests

To run tests, from the main pySAR folder run:

python -m unittest tests.MODULE_NAME -v

MODULE_NAME ->

Directory folders:

• /pySAR/PyBioMed - package partially forked from https://github.com/gadsbyfly/PyBioMed, used in the calculation of the protein descriptors.
• /Results - stores all calculated results that were generated for the research article, studying the SAR for a thermostability dataset.
• /pySAR/tests - unit and integration tests for pySAR.
• /pySAR/data - all required data and datasets are stored in this folder.

Contact

If you have any questions or comments, please contact: amckenna41@qub.ac.uk @

Back to top

|Logo| image:: https://raw.githubusercontent.com/pySAR/pySAR/master/pySAR.png

Install required dependencies and packages:

python setup.py install