MLProbs: A Data-Centric Pipeline for better Multiple Sequence Alignment

Evaluation

Run on all the benchmarks (in ./TEST/ folder):

cd ./path/to/MLProbs/
python script.py

Run on a single file:

cd ./path/to/MLProbs/
python MLProbs.py input.file.fasta [output.fasta.msa]

Installation

Please follow the README file in ./baseMSA/C_P_NP_Aln/ to compile our provided in ./baseMSA/ folder to c_p_np_aln and save it to ./baseMSA/C_P_NP_Aln/.

Please follow the quickprobs-manual.pdf to compile our provided QuickProbs in ./realign/QuickProbs/ to quickprobs and save it to ./realign/QuickProbs/bin/.

Use miniconda or pip to install sk-learn, subprocess, joblib, xlrd, xlwt or pip install -r requirements.txt to install them.

Main Dependency Libraries

PnpProbs: https://github.com/ytye/PnpProbs

QuickProbs: https://github.com/refresh-bio/QuickProbs

Scikit-Learn: https://scikit-learn.org/stable/ (recommend version: 0.21.3)

Others: subprocess, joblib, xlrd, xlwt (or just pip install -r requirements.txt)

Citation

@article{9705141,  
    author={Kuang, Mengmeng and Zhang, Yong and Lam, Tak-Wah and Ting, Hing-Fung},  
    journal={IEEE/ACM Transactions on Computational Biology and Bioinformatics},   
    title={MLProbs: A Data-Centric Pipeline for better Multiple Sequence Alignment},   
    year={2022},  
    doi={10.1109/TCBB.2022.3148382}
}