Strange EBSeq QC results

[Tian Yijun]

Hello everyone,

I am using rsem output (expected gene counts) to dissect some DE genes in EBSeq. I separate my data (GeneMat) into two group (n1=325; n2=142; Total Gene# 60448) and run EBSeq like:

SizesM=MedianNorm(GeneMat)

EBOut=EBTest(Data=GeneMat,Conditions=as.factor(rep(c('N8','GS8'), c(325,142)),sizeFactors=SizesM, maxround=15)

The convergence was like:

> EBOut$Alpha
           [,1]
iter1  2.540997
iter2  2.604404
iter3  2.593375
iter4  2.590792
iter5  2.583590
iter6  2.596551
iter7  2.589916
iter8  2.589916
iter9  2.589916
iter10 2.589916
iter11 2.589916
iter12 2.589916
iter13 2.589916
iter14 2.589916
iter15 2.589916

> EBOut$Beta
             Ng1
iter1  0.7084669
iter2  0.7097236
iter3  0.7073947
iter4  0.7080256
iter5  0.7058078
iter6  0.7095993
iter7  0.7082810
iter8  0.7082810
iter9  0.7082810
iter10 0.7082810
iter11 0.7082810
iter12 0.7082810
iter13 0.7082810
iter14 0.7082810
iter15 0.7082810

> EBOut$P
             [,1]
iter1  0.14271017
iter2  0.06926565
iter3  0.04553165
iter4  0.03786085
iter5  0.03444263
iter6  0.03302738
iter7  0.03248310
iter8  0.03248310
iter9  0.03248310
iter10 0.03248310
iter11 0.03248310
iter12 0.03248310
iter13 0.03248310
iter14 0.03248310
iter15 0.03248310

The QC plots were very strange (in the attachment). Can anyone help with this issue? Here is the link for QC plots.

https://drive.google.com/file/d/15b4xsSCYhyRfI8o_HOpdbuTdlCKJvkCQ/view?usp=sharing
https://drive.google.com/file/d/1kz3frHIZ4a6plS4410WkqfUYzbtuws7L/view?usp=sharing
https://drive.google.com/file/d/1zsTwERWFttR-n43HH8ZkNVM8gzDMYfBa/view?usp=sharing

Thank you!

Yijun