cyanxiao / covid-19-preprocess

Preprocess 🦠 COVID-19 strains data by the algorithm proposed by Tempel

COVID-19 Preprocess

Preprocess COVID-19 strains data by the algorithm proposed by Tempel. This is a relatively naive implementation of the original one aiming to test whether time-series training samples construction (explained in Tempel Fig.2) can be applied to COVID-19.

Data Files

src.paths.file_paths stores paths to the data which were not uploaded due to the file size limitation of GitHub. Missing data files are listed below:

• covid_amino_acids_path contains virus types and their corresponding amino acids sequence.
• covid_metadata_path contains virus types and their corresponding discovery time.
• covid_dna_sequences_path contains virus types and their corresponding DNA sequence.
• covid_protvec_path is the protVec of amino acids trigrams.
• covid_overlapping_trigrams_path contains the embedded sequences represented by numeric vectors.

Usage

Run src.execution.setup.