Sleuth_fit error "`Error in if (sum(valid) == 0) { : missing value where TRUE/FALSE needed"

Question

[williamsbrian5064]

Hi,

I am using Sleuth to generate some RNA-seq data. I keep getting this error when I try to run the program and I was hoping someone will be able to help me. When I try and run the so <- sleuth_fit(so, ~condition, 'full') I keep getting this error Error in if (sum(valid) == 0) { : missing value where TRUE/FALSE needed it looks like this is a common problem and it seems like an easy fix. Here are a couple link to people with similar issue example1 example2 example3. The so <- sleuth_prep(s2c, extra_bootstrap_summary = TRUE) works fine for me this is what it looks like

normalizing est_counts
23233 targets passed the filter
normalizing tpm
merging in metadata
summarizing bootstraps
..

I just don't know how they are fixing it. One person said "Thanks for the reply. And yes it appears to be a lack of replicates; when I just duplicated my dataset and fed it in as replicates it appears to work. Totally not a bug that the program doesn't run without replicates (what are you doing trying to perform differential expression analysis??!?), but if that appears to be the cause of this error, it might be nice to spit out a friendlier error ("Stop being an idiot and generate replicates!")." and another said "I re-ran my analysis once more with replicates and was able to run through sleuth_fit without any issues. The error I had last time was because my replicate samples were not labelled appropriately. For example, both replicates were labelled as Glucose_1 and Glucose_2 as opposed to just Glucose."

It appears it has to do something with running duplicates or replicates. I really don't know what they mean by that? I generated my data using kallisto. Do I have to do something different with kallisto or is this just a problem with sleuth? I know I'm doing the smallest thing wrong here I just can't figure it out.

Any help would be amazing!!

[Kevin Blighe]

You may have better luck posting on the GitHub page, where the developers of Kallisto / Sleuth appear to be more active. Alternatively , considering generating raw counts with Kallisto and then reading these into DESeq2 for downstream analyses.

[R]

Hello, how did you solve the problem?