contort 0.1.2

COnTROT (COmprehensive Transcriptomic ORganizational Tool) is a program that will download and organize all expression data in GEO related to a search result, commonly an organism.

COnTORT

Purpose:

COnTROT (COmprehensive Transcriptomic ORganizational Tool) is a program that will download and organize all expression data in GEO related to a search result, commonly an organism. This will identify and download GEO GDS, GSE, GSM, and GPL directories and files from NCBI FTP when provided with a downloaded GDS result from NCBI. Then the series data will be organized to retain matches to the GenBank annotation. The data will be organized, mean centered, and then joined using the gene name and ORF as IDs into a single text file that can be easily manipulated or opened in applications such as Excel.

To generate the initial file, perform a search in NCIB for GEO DataSets. Then in the upper right hand corner of the search results, select 'Send to:' and choose 'File' and then click 'Create File'. This will contain a summary of your search results. This script will search this file for series indicators and download the respective files and directories. You will also need the GenBank file for your organism.

COnTORT has been tested on Linux (Ubuntu and CentOS), MacOS (>10.13), and Windows 10.

Input :

• A gds_result text file downloaded from NCBI.

• The GenBank (.gbff) for your organism, downloaded from NCBI.

You are encouraged to create a new directory each time you run this script.

THIS WILL FAIL IF RUN ON VERSION OF PYTHON BELOW 3

Required Parameters:

-f gds_result.txt
-gb genebank_file.gbff

Usage:

contort -f gds_result.txt -gb genbank.gbff

Requirements:

• python 3
• R and R libraries GEOquery and tidyverse
• Python modles argparse, ftplib, functools, glob, gzip, io, os, pandas, re, shutil, subprocess, sys, time

To install the R libraries:

install.packages("tidyverse")

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install("GEOquery")

Output :

Primary output is a single text file containing the gene IDs on the first column and the mean centered expression data present in the remaining columns. The first row contains the headers references the specific GEO IDs for each experiment. Note that annotations no present in the GenBank file are listed with "N/A".

Locus_Tag    Old_Locus_Tag    Gene_Name    Gene_Synonyms    Product    GSM_ID_1    GSM_ID_2
RSP_0002     N/A              spbB         N/A              H-NS       12.0         4.0

Subdirectories are created for all the files and to organize the directory:

Directories created:

- geo                     - the downloaded GDS, GSE, GSM and GPL directories
- matrix_files            - the series_matrix.txt files downloaded from GEO
- GEOquery_results        - the results of the GEOquery.r script
- GFF_match_output        - the results of matching the GEOquery results to the gene IDs from the GFF
- mean_centered_results   - the mean centered expression data for each experiment
- FTP_files               - files used to download the data from GEO via FTP
- log_files               - all log files from each step as well as other saved files

Outline of steps & commands used in pipeline:

runFindGEOAddresses( GDSfile ):

Parse GEOfile

Opens and parses the GDSfile result txt file provided by the user (downloaded
from NCBI).

Creates new files:
    - GEO_FTP_Addresses.txt     - all GDS, GSE, GSM, GPL addresses in the file
    - GEO_FTP_directories.txt   - GDS, GSE, GSM, GPL directories that will be downloaded

run_is_ftp_dir( ftp_handle, name, guess_by_extension=True ):

QC for FTP

Determines if an item listed on the FTP server is a directory or not by 
looking for a "." in the fourth position. If it has that, it is nearly 
always a file and not a directory.

run_make_parent_dir( fpath ):

Make directories to match the FTP
Creates the directories in the local directory to match the FTP directories

run_download_FTP_file( ftp_handle, name, dest, overwrite ):

Copy the FTP files to the local directory
Copy FTP files into the respective directories on the local directory

run_mirror_ftp_dir( ftp_handle, name, overwrite, guess_by_extension ):

Replicate the directories
Replicates a direcotry from the FTP server onto the local drive recusively

download_ftp_tree( ftp_handle, path, destination, overwrite=False, guess_by_extension=True ):

Performs the actions to download files from the NCBI FTP

Perform the actions
Downloads an entire directory tree from an ftp server to the local destination
Will NOT overwrite files if present in the local directory

Default FTP Settings for NCBI GEO:

server = 'ftp.ncbi.nlm.nih.gov'
user = 'anonymous'
password = ''
destination = user input
sources of files = from the runFindGEOAddresses module

GEOquery( ):

Organize the GEO series files

This first finds and copies all series_matrix.txt.gz files from the FTP download.
These files are then unzipped and used to create the series.txt input file.
Run the Rscript GEO_annotate.r - requires GEOpatch.r
Runs GEOquery in R to download and combine the annotation with the log2 normalized
data in the GSE files.
This is then written to a new file for use in downstream steps.

gffMatch( GBFF ):

Match the gene annotations from the GenBank file to the organized expression data

Using dictionaries created from the GenBank file for the orgainsm, this
script will search for matches to the gene annotation in the GenBank file and
retain only those data with matches. This will make new files for each GEOquery
output with columns representing gene annotations and then the log2 normalized data
from the GEO series files. The data are then mean centered and joined together
with the gene annotations annotation as the key. All blanks are retained for consistency.
This file is written and can be used in Excel or R for further analysis.

cleanUp( cwd ):

Clean up the directory

Organize the files into folders for a cleaner directory