I need to try to compare some TCGA data with my rnaseq data. I take this 1098 sample and I try to use on deseq2 package.

SampleTable
     SampleName                                          fileName condition
1             1 0000772b-773d-4cf8-8baf-0e1e6dbf55e8.htseq.counts         A
2             2 003d640d-3bfc-4791-b241-4a3206d49468.htseq.counts         A
3             3 00511204-3512-4a5e-b664-60271e968903.htseq.counts         B
4             4 00587432-3004-4d4b-9dc2-8c90c81f7640.htseq.counts         B
5             5 0094a627-a55f-48dc-bbc3-ea66035e514a.htseq.counts         A
6             6 009ff7e3-8bd3-4683-adf5-7b68a0859f35.htseq.counts         A
7

This is the code I use. Unfortunately there is no way to obtain an Image. R kill the run when I try to use rlog()

Here you can se the code. What can I do for complete this task. My PC have 35 GB ram and Is Centos 7

metadata<-read.table("ID_provami.csv",sep=",",header=T)
SampleTable <-data.frame(SampleName=metadata$ID,fileName=metadata$NOMIFILE,condition=factor(metadata$CONDITION))

ddHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable=SampleTable,directory="TuttoCount", design= ~ 1) #

dds <- estimateSizeFactors(ddHTSeq)
#filtro 10 conti there are less than 0,3 samples with counts greater than or equal to 10. Given that the inference i
idx <- rowSums( counts(dds) >= 10 ) >= 0.3

dds <- dds[idx,]
dds <- DESeq(dds)

rld = rlog(dds)

data<-DESeq2::plotPCA(rld,intgroup=c("condition"),returnData=TRUE)
p1<-qplot(PC1,PC2,color=condition,data=data,size=I(9))+geom_text(aes(label=ID),hjust=-2,vjust=1)

ggsave("PCA_condition.png",p1)

Can you try vst()? I've worked on this function specifically to handle very large datasets. It also performs better on these datasets I believe.