Entering edit mode
Biomed
•
0
@biomed-17196
Last seen 6.2 years ago
Hi all,
It's my first-time to analysis Affymetrix HTA 2.0 arrays [transcript (gene) version], data source https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi. As a rookie, I have been bothered by its annotation for a long time, really hoping there would be someone teach me something. Thank you in advance!
The data probes like this: At first sight, there are 5 kinds of probes in this data？Sorry for my weak knowledge of probe sets.
| AFFX-r2-P1-cre-5_at |
| ERCC-00171_st |
| ERCCmix1step1 |
| ERCC-00002_st |
| JUC01000985.hg.1 |
Referring to some experience in Bioconductor, I used hta20transcriptcluster.db package to complete annotation, however, there are so many NA in the last result. I must have missed something. Are there someone so kind to help me understand my faults?
Thank you very much.
Thank you! Svlachavas.
Obviously, you are experienced in HTA 2.0. I have some other questions. Can you tell me the correct answer?
Q1, because the oligo can only support the pretreatment process of the oligonucleotide microarray, can I start the analysis directly with the GEO series Metrix files? I just want to filter the DEG. Another, I found that if I read the series Metrix files getGeo(), GEOquery packages, load data without probes (both are NA), of course, fData() is nothing. When I load another HTA 2.0 data, GSE93742 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi does not display this status. What are the potential reasons? Due to the strange situation above, I loaded the expression data after processing in Excel.
Q2, does the angle of view filter do use LIMMA? I read an article before that LIMMA can only filter probe level differences.
Looking forward to your reply!
Please, try to answer in English with a complete reply, i did not understand anything from your above answer !!
I'm so sorry for my careless! It's weird that I really add the reply in English! Apologize again!
Here is my formal answer:
Thank you! Svlachavas.
Obviously, you are experienced in HTA 2.0. I have some other questions. Can you tell me the correct answer?
Q1, because the oligo can only support the pretreatment process of the oligonucleotide microarray, can I start the analysis directly with the GEO series Metrix files? I just want to filter the DEG. Another, I found that if I read the series Metrix files getGeo(), GEOquery packages, load data without probes (both are NA), of course, fData() is nothing. When I load another HTA 2.0 data, GSE93742 https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi does not display this status. What are the potential reasons? Due to the strange situation above, I loaded the expression data after processing in Excel.
Q2, does the angle of view filter do use LIMMA? I read an article before that LIMMA can only filter probe level differences.
Looking forward to your reply!
Thanks for your precious time again.
Best wishes to you!
eSet=read.delim("GSE73219_series_matrix.txt")
昏暗（eSet）＃[1] 70523 13</font></font>
rownames(eSet)=eSet$ID_REF
eSet=eSet[,-1]
ESET=new("ExpressionSet", exprs=as.matrix(eSet))
##limma DEGs filtering
cell_line=factor(rep(c("Hep3B","Huh1"),each=6))
treat=factor(rep(c("scr","RDBP"),each=3,times=2))
combine=paste(cell_line,treat,sep="_")
design=model.matrix(~0+combine)
design=design[,c(2,1,4,3)]
colnames(design)=c("Hep3B_scr","Hep3B_RDBP","Huh1_scr","Huh1_RDBP")
rownames(design)=rownames(p_data)
cont_matrix=makeContrasts(Diff1=Hep3B_RDBP-Hep3B_scr, Diff2=Huh1_RDBP-Huh1_scr, DIFF=(Hep3B_RDBP-Hep3B_scr)-(Huh1_RDBP-Huh1_scr), levels = design)
fit1=lmFit(eSet,design)
fit2=contrasts.fit(fit1,cont_matrix)
result1=decideTests(fit2)
vennDiagram(vennCounts(result1))
result2=topTableF(fit3,number = Inf,p.value = 0.05)
fit3=eBayes(fit2)