PyMI-IPA

This repository contains a Python version of the Matlab MI-IPA implementation (GitHub): "Inferring interaction partners from protein sequences using mutual information", by Anne-Florence Bitbol (article).

---

• PyMI-IPA
  • What we know so far
    • Input
      • Standard_HKRR_dataset.fasta
      • SpeciesNumbering_Standard_HKRR_dataset.mat
    • Output
      • IniScrambling_Ninc1600_rep1.txt
      • Resf_Ninc1600_rep1.txt
      • TP_data_Ninc1600_rep1.txt
  • Contributing
    • Workflow
    • Setting up your local repository
    • Start coding
    • Dependencies

---

What we know so far (matlab version)

Input

Standard_HKRR_dataset.fasta

MSA in fasta format (5109 random pairs). First sequence is the reference, last a dummy sequence. Names look like this:

>Halhy_2679_HK_Classic|Haliscomenobacter_hydrossis_DSM_1100|Pair_Halhy_2678|HisKA_202-269/Halhy_2678_RR_unclassified|Haliscomenobacter_hydrossis_DSM_1100|Pair_Halhy_2679|Response_reg_6-118

>Halhy_3192_HK_Classic|Haliscomenobacter_hydrossis_DSM_1100|Pair_Halhy_3191|HisKA_204-270/Halhy_3191_RR_OmpR|Haliscomenobacter_hydrossis_DSM_1100|Pair_Halhy_3192|Response_reg_5-115

We should somehow input the info about the length of each sequence pair in the alignment (Probably best if given as an argument).

SpeciesNumbering_Standard_HKRR_dataset.mat

Sp. Index	Sp. name
1	Acetohalobium_arabaticum_DSM_5501
2	Carboxydothermus_hydrogenoformans_Z-2901
3	Vibrio_cholerae_MJ-1236
4	Rickettsia_rickettsii_str._Hino
5	Serratia_marcescens_WW4
6	Bacillus_cereus_AH820
7	Escherichia_coli_CFT073

502 Species in total. Made a .csv version for us.

Output

(from MatLab, see results.zip)

IniScrambling_Ninc1600_rep1.txt

Sp. Index	Seq nº	Sp. Index	Seq nº
491	4948	491	4949
491	4949	491	4944
491	4950	491	4947
491	4951	491	4936
492	4952	492	4955
492	4953	492	4956
492	4954	492	4954

Probably random pairs (5064), not final result.

Resf_Ninc1600_rep1.txt

Sp. Index	Seq nº	Seq nº	Score x?	Score y?
1	2	2	-822.35	452.64
1	3	3	-705.42	452.64
2	4	7	-769.9	136.96
2	5	5	-301.21	359.58
...	...	...	...	...
502	5105	5105	-605.18	385.49
502	5106	5108	-409.19	152.64
502	5107	5107	-285.82	457.96
502	5108	5098	-270.35	134.47

Contains 5064 lines with the final pairs (5109 pairs in the input fasta). We will have to match this seq. nº to the ones in the input fasta to generate a nicer output (table).

TP_data_Ninc1600_rep1.txt

Nº of pairs?	?	?	?	?	?
0	4452.3	872	4192	0	0
1600	1511	1455	3609	256	1344
3200	2240.5	1737	3327	1043	2157
4800	3014	1913	3151	1690	3110
5064	3156.3	2040	3024	1913	3151

Output matrix (MI_IPA_main.m: lines 102-104)

---

Contributing

Workflow

If you're interested in contributing to this project, please follow the following guidelines. Our main branch, where all changes will be gathered, is called "develop". Every "feature" (function) will have its own branch and once is finished, the "feature_branch" will be merged to "develop". The branch "matlab" contains the original matlab version. The pending tasks are managed within Issues.

Setting up your local repository

1. Clone the repository

git clone [email protected]:ElofssonLab/MI_IPA.git

or use the https version if you are not using SSH keys

git clone https://github.com/ElofssonLab/MI_IPA.git

1. Check current (local) branch

git branch

2. Check existing remote branches

git branch -r

3. Now you can collaborate in an existing branch or create a new one

• Create and switch to a new branch

git checkout -b <my_new_branch>

• Or download an existing branch ("MI_IPA_main.py" as an example)

git fetch origin MI_IPA_main.py:MI_IPA_main.py

Switch to the downloaded branch

git checkout MI_IPA_main.py

Start coding

Create the file ("my_code.py") with the code in the repository directory.
Go to the directory.

cd <path to directory>

Verify working branch

git branch

List files/directories modified locally

git status

Add to the stage your file

git add my_code.py

Commit the changes

git commit -m "Added my code"

Every commit creates a "snapshot" (GitHub advice about commits)

To finally send your commits to the remote:

git push

If is your first push to a newly created branch you will get this error:

fatal: The current branch my_new_branch has no upstream branch.
To push the current branch and set the remote as upstream, use
git push --set-upstream origin my_new_branch

So:

git push --set-upstream origin my_new_branch

---

Dependencies

• pandas
• biopython
• scipy
• munkres

Install with:

pip3 install pandas biopython scipy munkres