README - IGNORE(read internal READMEs)

This repository contains scripts for extracting Klebsiella pneumoniae data from CARD, creating a knowledge graph (KG), annotating Klebsiella pneumoniae genome sequence using a BAKTA/RGI pipeline, performing BLAST searches against a database, and filtering the BLAST results based on sequence identity.

Files and Functions

1. CARD/CARD_data2.py

Function:

• Extracts Klebsiella pneumoniae data from CARD and writes it into extracted_data_kleb.csv.

Modification:

• Modify line 43 to specify the desired bacteria for extraction. For this project, it extracts Klebsiella pneumoniae data.

Output:

• extracted_data_kleb.csv: A CSV file containing the data of Klebsiella pneumoniae.

2. CARD/create_kg.py

Function:

• Reads the extracted data from extracted_data_kleb.csv and creates a knowledge graph (KG).
• For each relationship between entities (e.g., drug targets, gene relationships), separate files are generated in the format of object-relationship-subject.

Output:

• Files representing various relationships in the knowledge graph.

3. DrugTarget/getTarget.py

Function:

• Reads UniProt IDs and antibiotic names from extracted_data_kleb.csv and uniprot_links.csv.
• For each UniProt ID:
  • Retrieves the corresponding protein sequence from UniProt in .fasta format.
  • Writes the sequence to query.faa.
  • Performs a BLAST search using the query sequence against the database (mergedsequenceskleb.faa).
  • Saves the BLAST results in blast_results.txt.

Functions Breakdown:

• get_protein_sequence(uniprot_id): Retrieves the protein sequence from UniProt for the given UniProt ID by making a request to the UniProt API.
• subprocess.run: Executes the BLAST command to compare the query sequence against the provided database. The results include details such as query accession, subject accession, e-value, query length, percentage identity (pident), query start/end, subject start/end.

Output:

• blast_results.txt: This file contains the BLAST search results for all the sequences compared to the database.

4. DrugTarget/identityFiltered.py

Function:

• Filters the BLAST results in blast_results.txt to retain only those hits where the percentage identity (pident) is greater than 90%.
• The filtered results are written to blast_hits_pident_above_90.txt.

Steps:

1. Reads the BLAST results from blast_results.txt.
2. For each hit, it checks if the pident is above the threshold (e.g., 90%).
3. If so, it saves the hit details (drug name, sequence, blast hit, and pident) to blast_hits_pident_above_90.txt.

Output:

• blast_hits_pident_above_90.txt: A filtered results file containing only BLAST hits with pident greater than 90%.

5. STRINGComparison/AMRsequences.py

Function:

• Extracts Klebsiella pneumoniae Gene sequences from CARD and writes it into query.faa. Uses the list of Genes related to Klebsiella pneumoniae from AMRGeneNames.csv

Modification:

• Modify line 7 to specify the desired output file.

Output:

• query.faa: A faa file containing the sequence data of each AMR Genes associated with Klebsiella pneumoniae.

6. STRINGComparison/filter.py

Function:

• Filters the BLAST results in specified file to retain only those hits where the percentage identity (pident) is greater than the specified threshold.

Input:

• Filter.py takes two command line inputs pident threshold and output file .

Output:

• filteredbypident.csv: A csv file containing the highest BLAST hits with threshold >=pident threshold.

7. annotation/geneNeighborhood.py

Function:

• Creates a gene network from the corresponding annotation_detail.csv file and embeds the network in a specified output text file.

Input:

• geneNeighborhood.py takes three command line inputs file_path to input annotation file, the output file path the network should be embedded in, and unkown_gene boolean if "True" to include unkownGenes in the gene Neighborhood and False exclude unkownGenes from the gene Neighborhood.

Output:

• geneNeighborhood.txt: A text file containing the gene Network derived from the genome sequence annotation.

Usage

1. Extract Data:

   • Run CARD/CARD_data2.py to extract Klebsiella pneumoniae data from CARD.
   • Ensure line 43 is modified to specify Klebsiella pneumoniae.

2. Create Knowledge Graph:

   • Use CARD/create_kg.py to read the extracted data from extracted_data_kleb.csv and create the knowledge graph files.

3. Retrieve Protein Sequences and Perform BLAST:

   • Execute DrugTarget/getTarget.py to read UniProt IDs and antibiotic names, retrieve protein sequences, perform BLAST searches, and save the results in blast_results.txt.

4. Filter BLAST Results:

   • Run DrugTarget/identityFiltered.py to filter the BLAST results based on percentage identity and save the filtered results to blast_hits_pident_above_90.txt.

5. Extract sequence Data:

   • Run STRINGComparison/AMRsequences.py to extract Klebsiella pneumoniae Gene sequences from CARD.json and write it into a specified output file.

6. Filter BLAST Results:

   • Run STRINGComparison/filter.py to filter the BLAST results based on percentage identity and save the filtered results to specified file name.

Dependencies

• Python 3.x
• Requests library for API calls
• BLAST+ for performing BLAST searches
• UniProt API for retrieving protein sequences

Ensure all dependencies are installed and properly configured before running the scripts.