会议纪要

  1. Add the part of NewRT to finish the application test on DIA18 and DIA185
    2. Model test for NewRT on PhosDIA three datasets to compare with DeepRT
    3. Optimize the conditions for virtual library generation to maximize the gained phospho-proteins, -sites, and -isomer groups. The stats of isomers contains both group level and modpep level.
    4. Add DIA185 as pretrain data after DDA and test on DIA18
    5. Training model with only DIA18 to compare with the model pre-trained with Jeff and DDA
    6. Additional model test for ion models with only the phospep isomers and compare the distribution with whole dataset
    7. How about the sequence with no label for pretrain. Directly add known phossites to protein sequence.

ToDo

1.Add the part of NewRT to finish the application test on DIA18 and DIA185
2.NewRT on PhosDIA three datasets to compare with DeepRT
3.Add DIA185 as training data, then test on DIA18 on model comparison and application level.
4.Evaluate NewSpec model on the phospep isomers.