Hi, I started playing with your clusterProfiler (btw, very nice idea

<div class="snippet-clipboard-content notranslate position-relative overflow-auto" data-snippet-clip

ERROR: The estimated pi0 <= 0... about clusterprofiler HOT 2 CLOSED

yulab-smu commented on June 29, 2024

ERROR: The estimated pi0 <= 0...

from clusterprofiler.

Comments (2)

GuangchuangYu commented on June 29, 2024

> xx=compareCluster(gene, "enrichKEGG")
[1] "ERROR: The estimated pi0 <= 0. Check that you have valid p-values or use another lambda method."
[1] "ERROR: The estimated pi0 <= 0. Check that you have valid p-values or use another lambda method."
[1] "ERROR: The estimated pi0 <= 0. Check that you have valid p-values or use another lambda method."
[1] "ERROR: The estimated pi0 <= 0. Check that you have valid p-values or use another lambda method."
[1] "ERROR: The estimated pi0 <= 0. Check that you have valid p-values or use another lambda method."
> xx
Result of Comparing 12 gene clusters
> summary(xx)
  Cluster       ID                             Description GeneRatio  BgRatio
1      V2 hsa04080 Neuroactive ligand-receptor interaction      8/10 272/5894
2      V3 hsa04080 Neuroactive ligand-receptor interaction     11/37 272/5894
3      V3 hsa04020               Calcium signaling pathway      7/37 177/5894
4      V3 hsa04144                             Endocytosis      6/37 203/5894
5      V5 hsa05200                      Pathways in cancer      6/16 327/5894
6      V7 hsa04080 Neuroactive ligand-receptor interaction      9/13 272/5894
7      V8 hsa04722          Neurotrophin signaling pathway      8/17 127/5894
8      V8 hsa04010                  MAPK signaling pathway      6/17 268/5894
        pvalue     p.adjust qvalue
1 7.728429e-10 7.728429e-10     NA
2 4.840797e-07 1.452239e-06     NA
3 9.395252e-05 1.409288e-04     NA
4 1.475332e-03 1.475332e-03     NA
5 1.388180e-04 1.388180e-04     NA
6 5.070086e-10 5.070086e-10     NA
7 7.699212e-10 1.539842e-09     NA
8 6.767807e-05 6.767807e-05     NA
                                             geneID Count
1           1813/1815/3062/135/1268/1812/11255/3061     8
2 4985/150/154/5021/153/2911/155/151/4986/3360/1128    11
3                   154/5021/153/2911/155/3360/1128     7
4                           408/154/156/409/153/155     6
5                     6257/6256/6258/5914/5294/5915     6
6      4160/4159/3953/3952/4889/4887/4886/2693/5739     9
7           4804/10782/4909/4915/4803/4916/627/4908     8
8                      5582/4909/4915/4803/627/4908     6

The error is due to the qvalue package can't estimate the qvalue properly. Although it throw error, clusterProfiler works just fine. Instead of qvalue, you can use the p.adjust values for multiple testing correction.

from clusterprofiler.