Skip to content
forked from nerve-bio/NERVE

NERVE is an user-friendly software environment for the in silico identification of the best vaccine candidates from whole proteomes of bacterial pathogens. The purpose of this project is to update it, implementing new modules with machine learning based methods, and improving the performance of the already implemented ones. Future directions of …

Notifications You must be signed in to change notification settings

FranceCosta/NERVE

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NERVE

Please visit https://nicolagulmini.github.io/NERVE/.

Stand-alone usage:

NERVE can be used as a stand alone verison taking advantage of Docker and Docker-compose in linux systems.

  1. install Docker following these instructions and the post-installation procedure
  2. install docker-compose as explained here
  3. clone repository:
git clone [email protected]:/FranceCosta/NERVE.git
  1. navigate to the correct folder:
cd NERVE
  1. Run NERVE (the first time it will take a few minutes)
./NERVE.sh --help

Expected output:

usage: NERVE.py [-h] [-a] [-ev] -g [-ml] [-mm] [-m] [-mpsl] -p1 [-p2]
                [-paefilter] [-pacfilter] [-pl] [-rz] [-rl] [-s] [-ss]
                [-tdl] [-vl] [-vir] [-wd] [-nd] [-id] [-dfd] [-ep]
                [-m1l] [-m2l] [-m1ovr] [-m2ovr] [-prt]

Run vaccine candidate prediction

optional arguments:
  -h, --help            show this help message and exit
  -a, --annotation  Activation (True) or deactivation (False) of annotation module.
                        Uses DeepFri to retrieve protein functional onthologies
                        (default: True)
  -ev, --e_value    Expect-value used in blastp for immunity modules (default:
                        1e-10)
  -g, --gram        Negative (n) or positive (p) gram stain of the pathogen of
                        interest (default: None)
  -ml, --minlength  Minimal length required for shared peptides to be extracted in
                        comparison analyses versus human and/or mouse (default: 9)
  -mm, --mismatch   Maximal number of not compatible substitutions allowed in shared
                        peptides alignment windows of 'minlength' size in immunity
                        modules (default: 1)
  -m, --mouse       Activation (True) or deactivation (False) of the mouse immunity
                        module. This module compares proteome1 with mouse proteome and a
                        further analysis of the eventual shared peptides is carried out
                        as in the autoimmunity module (default: True)
  -mpsl, --mouse_peptides_sum_limit 
                        Parameter calculated in mouse module and used by select module.
                        Protein with 'sum of shared peptides of the i-protein with mouse
                        proteins/number of aminoacids of the i-protein' <=
                        mouse_peptides_sum_limit and with absence of match mhc-I and
                        Mhc-II mouse ligands are selected (default: 0.15)
  -p1, --proteome1  Path to proteome or Uniprot proteome ID (see:
                        https://www.uniprot.org/proteomes/?query=&sort=score) (default:
                        None)
  -p2, --proteome2  Path to proteome or Uniprot proteome ID (see:
                        https://www.uniprot.org/proteomes/?query=&sort=score) (default:
                        None)
  -paefilter, --p_ad_extracellular_filter 
                        Parameter of select module. Extracellular proteins with a
                        probability of adhesin (pad) lower than
                        p_ad_extracellular_filter are discarded (0.-1) (default: 0.38)
  -pacfilter, --p_ad_no_citoplasm_filter 
                        Parameter of select module. Non-cytoplasmic Proteins with a
                        probability of adhesin (pad) lower than p_ad_no_citoplasm_filter
                        are discarded (0.-1) (default: 0.46)
  -pl, --padlimit   Set the probability of adhesin (pad) value cut-off for proteins
                        with 'Unknown' localization in the select module. Thus, these
                        proteins with a pad value < cut-off are discarded (0.-1)
                        (default: 0.85)
  -rz, --razor      Activation (True) or deactivation (False) of the loop-razor
                        module. This module allows the recovery of protein vaccine
                        candidates, with more than 2 transmembrane domains, that would
                        otherwise be discarded in the select module. The longest loop
                        with minimum len == 'razlen' aa will replace the original
                        protein sequence for following NERVE steps (default: True)
  -rl, --razlen     Set minimal length of loop considered in loop-razor module
                        (default: 50)
  -s, --select      Activation (True) or deactivation (False) of select module,
                        which filters PVC from proteome1 (default: True)
  -ss, --substitution 
                        Maximal number of compatible substitutions allowed in shared
                        peptides alignment windows of 'minlength' size in immunity
                        modules (default: 3)
  -tdl, --transmemb_doms_limit 
                        Parameter of select module. Proiteins with trasmembrane domains
                        >= transmemb_doms_limit are discarded (default: 3)
  -vl, --virlimit   Cut-off value for NERVirulent in the select module (0.-1)
                        (default: 0.5)
  -vir, --virulent  Activation (True) or deactivation (False) of NERVirulent module,
                        predictor of the probability of being a virulence factor
                        (default: True)
  -wd, --working_dir 
                        Path to working directory. If not existing, a working directory
                        with the given path is created (default: ./)
  -nd, --NERVE_dir  Path to NERVE repository folder (download from:
                        https://github.com/nicolagulmini/NERVE) (default:
                        /usr/nerve_python/NERVE)
  -id, --iFeature_dir 
                        Path to iFeature repository folder (download from:
                        https://github.com/Superzchen/iFeature) (default:
                        /usr/nerve_python/assets/iFeature)
  -dfd, --DeepFri_dir 
                        Path to DeepFri folder (download from:
                        https://github.com/flatironinstitute/DeepFRI) (default:
                        /usr/nerve_python/assets/DeepFri)
-ep (epitope prediction module), [True, False, default=True];
-m1l (mhc1 ligands length), [9,10,11, default=9];
-m2l (mhc2 ligands length), [9,11,13,15, default=11];
-m1ovr (mhc1 ligands max overlap), [1,2,default=1];
-m2ovr (mhc2 ligands max overlap), [1,2, default=1];
-prt (epitope binders percentile), [float, default=0.80]

Example run:

./NERVE.sh -p1 UP000002493 -g n -wd ./UP000002493

About

NERVE is an user-friendly software environment for the in silico identification of the best vaccine candidates from whole proteomes of bacterial pathogens. The purpose of this project is to update it, implementing new modules with machine learning based methods, and improving the performance of the already implemented ones. Future directions of …

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 97.4%
  • Python 2.3%
  • Other 0.3%