transcript and exon level analysis for detection of alternative splicings

37 views
Skip to first unread message

jjspring OH

unread,
Sep 21, 2014, 9:50:25 PM9/21/14
to aroma-af...@googlegroups.com


Hi,


I would like to compare the results from plmTr and plmEx, estimation of overall expression for transcripts and exon-by-exon estimation,
but when i tried to run ExonRmaPlm with mergeGroups = T and F, respectively, i am getting the identical results for following objects including FIRMA scores based on 29170 rows by samples,
 


plmTr <- ExonRmaPlm(csN, mergeGroups=TRUE)

print(plmTr)



plmEx <- ExonRmaPlm(csN, mergeGroups=FALSE)

print(plmEx)

...



firma <- FirmaModel(plmEx)

fit(firma, verbose=verbose)

fs <- getFirmaScores(firma)

asData <- extractDataFrame(fs, addNames=TRUE)


> fs <- getFirmaScores(firma)

> asData <- extractDataFrame(fs, addNames=TRUE)

> dim(asData)

[1] 29170    13





Yet, And also, when i ran previously in oligo library, got 213067 probe indices 



library(oligo)


celFiles = list.celfiles('/Users/sungheeoh/Desktop/SUNGHEE/PROJS/KANG/RAW-CELFILES/',full.name =T)

raw = read.celfiles(celFiles)

exon_rma = rma(raw,target="probeset")  #pre-processing/normalization/log2-scale

exon_mat <- exprs(exon_rma)

dim(exon_mat)

head(exon_mat)



> dim(exon_mat)

[1] 213067      8


My questions,


first question is how i can get different sets from transcript and exon-by-exon from cell files?


and also, just curious, how is it mapped from probe sets (213K) onto transcript/exon level annotation(29K)?  


And, major interest is to detect differentially alternative splicings for each gene? (we are mostly interested in looking at those genes, e.g. Although a gene is differentially expressed but it is not differentially expressed in AS or vice versa,) how can i extract these information from FIRMA scores? 



could you pls let me know ?


Thanks a lot in advance, 


Sunghee




jjspring OH

unread,
Sep 21, 2014, 10:19:09 PM9/21/14
to aroma-af...@googlegroups.com



And one more question:

On the FIRMA scores, what are NA values, though expression levels are not zero?


Thanks, 


Sunghee



2014년 9월 22일 월요일 오전 10시 50분 25초 UTC+9, jjspring OH 님의 말:

Sanjeev Sariya

unread,
Apr 17, 2018, 2:20:18 PM4/17/18
to aroma.affymetrix
Hi Sunghee,

Did you figure this out? I'm in the same boat.
Reply all
Reply to author
Forward
0 new messages