mave2imap 1.0.0.0.4

This project is intended for infering and mapping interface hotspots based on results from MAVE (Multiplexed Assays of Variant Effects).

mave2imap

Table of contents

• mave2imap
• • Table of contents
  • Description
  • Install (Linux)
  • Testing
    • 1) Create a folder to download required data and run the test :construction:
    • 2) Run mave2imap pipeline for each targeted region. :computer:
    • 3) Analyze results using jupyter notebook(s). :mag_right:
  • Citing mave2imap
  • Copyright
    • Acknowledgements

---

Description

This code is intended for 3D mapping of interface hotspots based on the most perturbed positions inferred from MAVE (Multiplexed Assays of Variant Effects) results. (See publication)

---

Install (Linux)

$ conda env create -f https://github.com/synth-bio-evo/mave2imap/blob/main/mave2imap.yml

---

Testing

Requires about >= 64 Gb RAM to process the full dataset.
If you do not dispose of this amount of RAM you can create smaller .fastq.gz files by using the following command:

gunzip -cd <file>.fastq.gz | head -n 1600000 | gzip > <file_400k_reads>.fastq.gz

• Replace "<file>" by your filename
• It will extract and compress 1,6x10⁶ lines from "<file>.fastq.gz", corresponding to 4x10⁵ reads, and create "<file_400k_reads>.fastq.gz"

1) Create a folder to download required data and run the test :construction:

mkdir /tmp/test
cd /tmp/test

If you have aria2c installed (faster)

aria2c -j 16 <link>

Else

wget <link>

Uncompress the .tar.gz file

tar -xvzf Asf1B+IP3.tar.gz

2) Run mave2imap pipeline for each targeted region. :computer:

Exemple:

cd Asf1B+IP3/Asf1_N-Ter
mave2imap -i Asf1_N-ter.ini
cd ../Asf1_C-Ter
mave2imap -i Asf1_C-ter.ini

This will produce the data required for analysis and visualization using the proposed jupyter notebook.

:microscope: The information available in the output file, "result_thresh3_2_2_compare_conditions.out", is probably the most relevant to a classical user.

3) Analyze results using jupyter notebook(s). :mag_right:

• enter appropriate folder and launch jupyter-lab
  
  

for interface mapping:

cd ../imap_notebook

for fitness assessement:

cd ../fitness_notebook

• for both

jupyter-lab

• Choose mave2imap kernel
• If required edit the code according to your specific case (not required for the testing dataset)
• Click in "Run" (menu) => "Restart Kernel and Run All Cells"

The most perturbed positions should be indicated below the last cell based on the defined threshold and you should be able to visualized/manipulated the 3D interactive complex (most perturbed regions are indicated by reddish gradient)

---

Citing mave2imap

"Publication is coming ..."

---

Copyright

Copyright (c) 2025, Raphaël Guérois (CEA-Saclay/DRF/Joliot/I2BC/SB2SM/LBSR), Oscar H.P. Ramos (CEA-Saclay/DRF/Joliot/MTS/SIMoS/LICB/SBE)

Acknowledgements

Project based on the Computational Molecular Science Python Cookiecutter version 1.11.