Open sgosline opened 2 years ago
We need to determine if the imputation can help our data.
I'd first run the DREAM AI tool, which has instructions here: https://github.com/PNNL-CompBio/proteomicsTumorDeconv/tree/main/imputation
Let's run this about 10 times to create 10 imputed datasets, then plot PCA of each of the datasets side by side, compared to the original data. If things look stable, we can continue.
Pretty futile :(
CamiloPosso
commented
2 years ago
We need to determine if the imputation can help our data.
I'd first run the DREAM AI tool, which has instructions here: https://github.com/PNNL-CompBio/proteomicsTumorDeconv/tree/main/imputation
Let's run this about 10 times to create 10 imputed datasets, then plot PCA of each of the datasets side by side, compared to the original data. If things look stable, we can continue.