furacca / brivez

Brivez is a bioinformatic tool thought as Quality of Life's improvement, providing high quantity of data in a snap, giving you a quick view on what you could find inside your transcriptome/sequences' list.

Brivez is a bioinformatic tool thought as Quality of Life's improvement, providing high quantity of data in a snap, giving you a quick view on what you could find inside your transcriptome/sequences' list.

Installation requirements aside, you will need two files:

• domain_profile.hmm
  It goes inside --> 01_hmm_profiles folder
• sequence(s).fasta
  It goes inside --> 02_fasta_target

Suggested way to proceed:

1. Read the documentation
2. The first time follow the checklist
3. Run ./brivez_main.sh

At the moment this program runs exclusively on Linux (tested on Debian 11 and Ubuntu 22.04).

See future updates for Mac and Windows.

All the software used are OpenSource.

Brivez runs locally
Multi-core CPU is suggested Total space used: ~3.5 GB SSD is strongly suggested 16 GB RAM is suggested

Index

• Workflow
  
• Software requirements
  
• Quickly set up
  
• First run
  
• Read results
  
• Useful link
  
• Future updates
  

Workflow

Brivez analyses every fasta file insiede 02_fasta_target.
Every single run follow these workflow:

• Runs Deepsig looking for Signal Peptide (can be disabled with -a)
• Uses the sequences found in the previously step
• Searches for domains using the hmm file inside 01_hmm_profiles
  • If Brivez founds more than one hmm file, it asks which one choose
• Generates multiple output files for each fasta target
  • A file with all the domain extraced can be genereted with -b option

Flowchart realized with to draw.io_

Software requirements

• Conda (minimum ~3 GB)
• Environment inside Conda with:
  • Deepsig (~50 MB) --> needs PYTHON 3.8 !
  • Pandas (~15 MB)
  • Bio-conda (channel)
  • fnmatch (samtools ~1 MB)
• HMMER3 v3.3.2 (~20 MB)
• Brivez (~1 MB)

Quickly set up

01 - Quick install for Conda (following the online doc is suggested):

1. Download the installer at this link
2. Verify your installer hashes
3. In the terminal run
   bash Anaconda-latest-Linux-x86_64.sh
   Create an environment ad hoc
   conda create -n bioinfo-brivez python=3.8
   To see all the env created
   conda env list
4. Choose the environment with
   conda activate bioinfo-brivez
   to close it use
   conda deactivate bioinfo-brivez

---- By this point be sure to have activated the right environment! ----

02 - Install in Conda some stuff:

• conda config --add channels bioconda
  
• conda install -c bioconda samtools
  
• conda install -c conda-forge dpath
  
• conda install pandas

03 - Install the predictor of signal peptides, DeepSig

• Just as described on its site, use
  conda install -c bioconda deepsig

04 - Install HMMER3

• Just as described on its site, use
  sudo apt-get install hmmer
  (v3.3.2 both on Ubuntu 22.04 and Debian 11)
  Otherwise is installable following the official documentation

05 - Download Brivez

• Download the Brivez folder with
  git clone https://github.com/furacca/brivez
  or whatever way you prefer

First run

1. 01_hmm_profiles

   • contains your hmm files (check hmm - Pfam download)

2. 02_fasta_target

   • contains your fasta files
   • the only accepted format is fasta (no fas, fa, ...)

3. Be sure to have the right environment activated
   

   • conda activate therightenvironment

4. Run Brivez with
   ./brivez_main.sh

Read results

Inside the 03_results you will have your output folder, all with the same structure:

• Research_number_XXX
  • fasta_target1 (directory)
    • multiples files
  • fasta targetX (directory)
    • multiples files
  • Report (file)

For every fasta target there will be one folder containings the following:

• 01_deepsig-analysis.tsv
  • Deepsig analysis; if -a is enabled, then is empty
• 02_sequences_selected.fasta
  • Sequences with SP by deepsig; if -a is enabled, all sequences are selected
• 03_hmmsearch_output_table
  • hmmsearch output
• 04_hmmsearch_output_table_data_parsed
  • hmmsearch output parsed in csv
• 05_pandas_sequences_table.tsv
  • pandas table with the list of all domains found
• 06_sequences_which_have_domains.fasta
  • list of sequences presenting domains found
• 07_extracted_domains.fasta
  • domains found extracted by the selected sequences
• 08_extracted_domains.csv
  • same as 07, but in csv format
• 09_domains_found.csv

Important The 08_extracted_domains.csv file can be easyly read in calc/excel selecting all the cells and clicking on Data -> Autofilters.

Useful link

• hmm - Pfam download
  Contains tutorials on:
  • How to download the full Pfam database
  • How to download a specific ~.hmm file from pfam
• Hmmer cheatsheet

Known problems

None (yet).

Future updates

TOP PRIORITY

• None

MEDIUM PRIORITY

• Simpler way to user Brivez (Docker? Appimage?)
• List of the sequences with just one domain type (needing full Pfam database)

LOW PRIORITY

• Check some solution for Mac and Windows (Docker?)
• Extracted list of all SP