Secondary structure analysis research: formal grammars + neural networks

Solution structure

Parsing tool

• Compiled version
• Source code

Requirements (Ubuntu)

• Install mono
• Install .net framework 4.7.2 via wine
• Install dotnet-sdk-3.0
• dotnet add package FSharp.Core --version 4.5.0

Run

./YaccConstructor/src/SecondaryStructureExtracter/bin/Debug/SecondaryStructureExtracter.exe -g 'grammar.txt' -i 'input_sequences.fasta' -o 'output_dir/'

Additional argument: output file format -f csv (default — bmp)

Grammar example is presented in the repo root folder

Neural networks

Requirements

• Python 3
• TensorFlow-gpu 1x
• Keras, scikit-image

Repository description

Experiments

• 16S detection — binary classifier that separates true 16s RNA sequences from random parts of genome
• Chimeras detection — not comleted research for chimeric sequences search in biological databases
• Secondary structure prediction — model for predicting RNA secondary structure contact map from parsing provided matrix
• TRNA classification — solutions for several trna classification tasks with small amount of classes

Content

• Links to all required datasets and trained models weights are presented in data/datalinks.txt files
• Models code and logs are presented in model/ folder
• All useful data processing scripts are stored in scripts/ folder

Papers

• Grigorev S., Lunina P. The Composition of Dense Neural Networks and Formal Grammars for Secondary Structure Analysis //BIOINFORMATICS. – 2019. – С. 234-241.
• Grigorev S., Lunina P. Improved Architecture of Artificial Neural Network for Secondary Structure Analysis //BMC Bioinformatics. – 2019. – Т. 20. – №. S17. – С. P2.
• Lunina P., Grigorev S. On Secondary Structure Analysis by Using Formal Grammars and Artificial Neural Networks //International Meeting on Computational Intelligence Methods for Bioinformatics and Biostatistics. – Springer, Cham, 2019. – С. 193-203.
• Grigorev S., Kutlenkov D., Lunina P. Secondary structure prediction by combination of formal grammars and neural networks //BMC Bioinformatics. – 2020. – Т. 21. – №. SUPPL 20.
• Комбинирование нейронных сетей и синтаксического анализа для предсказания вторичных структур генетических цепочек
• Комбинирование нейронных сетей и синтаксического анализа для обработки вторичной структуры последовательностей
• Комбинирование нейронных сетей и синтаксического анализа для предсказания вторичной структуры РНК
• Genegram: RNA Secondary Structure Prediction Using Formal Grammars and Residual Neural Networks

---

Contact: lunina_polina@mail.ru