Hello all, I have the FPKM values from NORMAL vs TUMOR.I am doing GSEA analysis using GSEA guy.
Below is my input（lncRNA_up.gct）:
1.2
130 100 NAME DESCRIPTION SRR934732.sorted.bam SRR934781.sorted.bam RP3-340N1.5 na 0.405482833 0.709856841 AC116614.1 na 0.583072119 0.777716411 XLOC_000564 na 523.3443013 53.65888594 …… I have a gmt format file from GSEA(c2.cp.kegg.v6.0.symbols.gmt).When I try to Run GSEA,I get the following error: 4891 [INFO ] Begun importing: Dataset from: lncRNA_up.gct
of elements = 1
lncRNA_up.cls 5032 [INFO ] Begun importing: Template from: lncRNA_up.cls 5049 [INFO ] No dataset collapsing was done .. using original as is to parse>c2.cp.kegg.v6.0.symbols.gmt< got: [c2.cp.kegg.v6.0.symbols.gmt] 5049 [INFO ] Begun importing: GeneSetMatrix from: c2.cp.kegg.v6.0.symbols.gmt 5083 [INFO ] Got gsets: 186 now preprocessing them ... min: 15 max: 500 5101 [INFO ] Done preproc for smaller than: 15 xtools.api.param.BadParamException: After pruning, none of the gene sets passed size thresholds. at xtools.api.param.ParamFactory.checkAndBarfIfZeroSets(ParamFactory.java:594) at xtools.gsea.Gsea.execute(Gsea.java:109) at xtools.api.AbstractTool.tool_main(AbstractTool.java:620) at xtools.gsea.Gsea.main(Gsea.java:136) 5111 [INFO ] Renaming rpt dir on error to: /mnt/pub/work/liuzhe/projs-cas-ibp/xuelab/GSCs_lncRNA/GSEA_analysis/Up-reagulated/apr07/error_my_analysis.Gsea.1491526934954
My parameters are: java -Xmx1024m -cp /mnt/pub/work/liuzhe/tools/gsea2-2.2.4.jar xtools.gsea.Gsea -res lncRNA_up.gct -cls lncRNA_up.cls -gmx c2.cp.kegg.v6.0.symbols.gmt -collapse false
Has anyone encountered similar Error? Thank you very much!
I've used it previously for microarray data(form GSEA example dataset P53) and it worked fine.
It seems that the filters to select the gene sets are too stringent, and you end up without any gene set to test with. Try increasing the limits of genes in gene sets. 5 in the lower band and 5000 in the upper band
Thank you! From your comment, I known add a parameter min size and/or max size can overcome the question. Thanks for your kindly help!