- data: human genome, chimpanzee genome, dog genome
- learned how to working with DNA Seqeunce, 3 way to encode genome sequence.
- train a machine learning model with human genome
- percision, recall, F1 socre were used as accuracy metric.
- data: wheat genotype, and wheat yield form different field
- calculating SNP signifincant
- lasso regression used as bench model
- 3 layers neural network and 4 layer CNN.
- pearson correlation was used as accuracy.
- inspired by the paper and reference
- "https://www.researchgate.net/publication/334609713_A_Guide_for_Using_Deep_Learning_for_Complex_Trait_Genomic_Prediction"