Getting error when creating an idea object

[Sunyp-IM]

I followed the instruction in the tutorial to analyze my data. After run the CreateiDEAObject command, I got the following output.

  > idea <- iDEA.fit(idea,
+                  fit_noGS=FALSE,
+                  init_beta=NULL, 
+                  init_tau=c(-2,0.5),
+                  min_degene=5,
+                  em_iter=15,
+                  mcmc_iter=1000, 
+                  fit.tol=1e-5,
+                  modelVariant = F,
+                  verbose=TRUE)
## ===== iDEA INPUT SUMMARY ==== ##
## number of annotations:  100 
## number of genes:  1384 
## number of cores:  1 
## fitting the model with gene sets information... 

However, looking at the result of idea@gsea, all the 100 annotations have pvalues >0.5. I wonder how to test more annotations and find the most significant ones for my gene set.

Thank you!

[YingMa0107]

Hi @Sunyp-IM,

Thank you for your interest in our package. Sorry for the late reply as I somehow cannot receive the email notification.

Based on the iDEA input summary, the total number of genes you provided in the summary stats is too small. iDEA requires all gene levels summary statistics not just DE genes. Could you please tell me how did you get the summary statistics? Did you only maintain the DE summary statistics as input for iDEA? If so, please use all gene levels summary statistics.

Best, Ying

[denvercal1234GitHub]

Hi @YingMa0107 - I hope to get clarification regarding the input to iDEA.

In the tutorial, it says "Assume you have obtained the DE results from i.e. zingeR, edgeR or MAST with the data frame res_DE (column: pvalue and LogFC)", so it does seem that the input can be output of DEG analysis such as via Seurat::FindAllMarkers(). In this case, do I need to create 1 iDEA object for each cluster individually (i.e., 1 cluster's DEG list as 1 iDEA object)? Is it more preferable to do for all clusters as 1 object?

Thank you for your help!