Question

clusterProfiler setup specified order

0

Entering edit mode

joseph ▴ 50

@joseph-5658

Last seen 7.1 years ago

I have time-course data as "Day1", "Day2"....., when I create the list for compareCluster(), I have named the order as levels(list) = c("Day1","Day2"....), and I check the compareCluster() result, CompGO@compareClusterResult$Cluster, shows the order is "Day1", "Day2".... However, when I plot the data, the X-axis show the disorder characters. How can I fix the problem?

clusterProfiler • 3.3k views

ADD COMMENT • link 7.7 years ago joseph ▴ 50

0

Entering edit mode

show me the code

ADD REPLY • link 7.7 years ago Guangchuang Yu ★ 1.2k

0

Entering edit mode

list = list(Day1 = Day1, Day2 = Day2, Day3 = Day3......)

levels(list) = c("Day1","Day2","Day3"....)

lx = compareCluster(list)

dotplot(lx)

ADD REPLY • link 7.7 years ago joseph ▴ 50

0

Entering edit mode

already fixed, <https://github.com/GuangchuangYu/clusterProfiler/commit/7fda4a36409a112c5b994dbbef47483aa3929f9a>.

The order in input list will be maintained when plotting.

ADD REPLY • link 7.7 years ago Guangchuang Yu ★ 1.2k

0

Entering edit mode

should I update to the latest version? I install the ver-3.4.1. The one you fixed in your developed version?

BTW, after compareCluster(), might be some time point not have the gene clusters, but in order to present the time-course data, can I keep show up all time points data, even if some of them haven't significant enriched pathway/GO?

ADD REPLY • link 7.7 years ago joseph ▴ 50

0

Entering edit mode

Could you please clarify how to place the specific order on the x-axis? And keep some groups even they don't have enriched result, just to show the longitudinal data,

ADD REPLY • link 7.7 years ago joseph ▴ 50

score 0 · Answer 1 · 2017-04-30

0

Entering edit mode

joseph ▴ 50

@joseph-5658

Last seen 7.1 years ago

And one more problem when I use enricher() function, also return no gene mapped:

egmt_go <- enricher(DEG_Day9_ENTREZID, TERM2GENE = c_go,
+ pvalueCutoff = 0.5, pAdjustMethod = "BH", universe,
+ minGSSize = 10, maxGSSize = 500, qvalueCutoff = 0.2)
--> No gene can be mapped....
--> Expected input gene ID: DNAJB1,ZFP36,PDE3A,POLE,AADAT,DIS3
--> return NULL...

The genes have been converted into EntrezID, more than 500 gene, should be match to GO or KEGG database.

ADD COMMENT • link 7.7 years ago joseph ▴ 50

0

Entering edit mode

your c_go contains gene symbol while DEG_Day9_ENTREZID is entrezgene ID.

ADD REPLY • link 7.7 years ago Guangchuang Yu ★ 1.2k

0

Entering edit mode

I use entrezID for both of them, but has the error as

> egmt_kegg <- enricher(DEG_Day9_ENTREZID, TERM2GENE = c_kegg,
+ pvalueCutoff = 0.1, pAdjustMethod = "BH", universe,
+ minGSSize = 10, maxGSSize = 500, qvalueCutoff = 0.2)
Error in as.vector(x, mode) :
cannot coerce type 'closure' to vector of type 'any'

ADD REPLY • link 7.7 years ago joseph ▴ 50

0

Entering edit mode

you need to provide a reproducible example as described in <http://guangchuangyu.github.io/2016/07/how-to-bug-author/>.

ADD REPLY • link 7.6 years ago Guangchuang Yu ★ 1.2k