statbiophys / sonia

Infer selection pressures on features of amino acid CDR3 sequences.

SONIA

This Package is not actively mantained, please refer to the most updated SoNNia package which includes sonia models as well: https://github.com/statbiophys/soNNia

SONIA is a python 3.6/2.7 software developed to infer selection pressures on features of amino acid CDR3 sequences. The inference is based on maximizing the likelihood of observing a selected data sample given a representative pre-selected sample. This method was first used in Elhanati et al (2014) to study thymic selection. For this purpose, the pre-selected sample can be generated internally using the OLGA software package, but SONIA allows it also to be supplied externally, in the same way the data sample is provided.

SONIA takes as input TCR CDR3 amino acid sequences, with or without per sequence lists of possible V and J genes suspected to be used in the recombination process for this sequence. Its output is selection factors for each amino acid ,(relative) position , CDR3 length combinations, and also for each V and J gene choice. These selection factors can be used to calculate sequence level selection factors which indicate how more or less represented this sequence would be in the selected pool as compared to the the pre-selected pool. These in turn could be used to calculate the probability to observe any sequence after selection and sample from the selected repertoire.

Version

Latest released version: 0.2.2

Installation

SONIA is a python 2.7/3.6 software. It is available on PyPI and can be downloaded and installed through pip:

pip install sonia.

SONIA is also available on GitHub.

Sometimes pip fails to install the dependencies correctly. Thus, if you get any error try first to install the dependencies separately:

pip install tensorflow
pip install matplotlib
pip install olga
pip install sonia 

For mac user on new metal devices, make sure to install additional dependencies. Currently, the configuration tensorflow-macos==2.9 and tensorflow-metal==0.5.0 should work.

References

1. Sethna Z, Isacchini G, Dupic T, Mora T, Walczak AM, Elhanati Y, Population variability in the generation and thymic selection of T-cell repertoires, (2020) bioRxiv, https://doi.org/10.1101/2020.01.08.899682
2. Isacchini G, Sethna Z, Elhanati Y ,Nourmohammad A, Mora T, Walczak AM, Generative models of T-cell receptor sequences, (2020) Phys. Rev. E 101, 062414, https://journals.aps.org/pre/abstract/10.1103/PhysRevE.101.062414
3. Elhanati Y, Murugan A , Callan CGJ , Mora T , Walczak AM, Quantifying selection in immune receptor repertoires, PNAS July 8, 2014 111 (27) 9875-9880, https://doi.org/10.1073/pnas.1409572111

Documentation

Extensive documentation can be found here

Contact

Any issues or questions should be addressed to us.

License

Free use of SONIA is granted under the terms of the GNU General Public License version 3 (GPLv3).

sonia's Issues

Main task list

• Stream code of update_model
• Save/Load to include features (Binary?)

Sequence generation failing on Windows machine

Hi, I was exploring your modules for sequence generation. I stumbled across one issue appearing in sequence_generation.py and sonia.py scripts. I was getting below error:

ValueError: high is out of bounds for int32

for the part of code that is generating the seeds values:

seeds=np.random.randint(2**32-1,size=num_gen_seqs)

Since no dtype is provided for np.random.randint method, numpy is using a default int, which is platform dependant. For Windows64 machines this will be int32 (you can read about it here: https://stackoverflow.com/questions/36278590/numpy-array-dtype-is-coming-as-int32-by-default-in-a-windows-10-64-bit-machine). Since it's a signed 32 bits integer, the 2**32-1 value is out of bounds. You can easily fix it by specifying dtype while generating seeds, for example:
seeds=np.random.randint(2**32-1,size=num_seqs,dtype=np.int64)