Question on Cuffdiff’s DE results.
Code:
Quote:
cuffdiff -p $NSLOTS -o cuffdiff_v110 Ensembl_59.gtf C_3732_hits.bam,C_6341_hits.bam,C_6356_hits.ba
m,C_6124_hits.bam,C_60PO_hits.bam HD_4430_hits.bam,HD_4412_hits.bam,HD_4242_hits.bam,HD_4189_hits
.bam,HD_3584_hits.bam
Versions
v091 v101 v102 v103 v110
OK
18112 33215 33230 33216 6765
Fail
3957 2524 2509 2523 1234
LOWDATA
NA 514 514 514 9000
NOTEST
29582 15408 15408 15408 34634
Total
51651 51661 51661 51661 51633
We wouldn’t expect such a big difference[1] between v110 and v101(v102,v103)especially in the numbers of NOTEST and OK. Cole Trapnell explained what does “Fail” mean [2], we would like to know what “NOTEST” means here.
All default options for the two versions appear to be the same; the command line options are exactly the same when we use different versions.
The cuffdiff v1.1.0 indicated.
“Cuffdiff now includes a more sophisticated check for sufficient sequencing depth prior to testingfor differences, which substantially improves the accuracy of differential expression analysis in loci with low to medium depth.”
Question on Cuffdiff’s q-value.
What is q-value here? Lots of evidence suggests q-value should be less than one [4], [5].
Reference:
(1) Inconsistency between cuffdiff 1.1.0 and cuffdiff 1.0.2
(2) Cole Trapnell on “Fail” test
(3) Cuffdiff producing q-values > 1
(4) Understanding q-values as a More Intuitive Alternative to p-values
(5) P-values, False Discovery Rate (FDR) and q-values What are p-values?
(6) Cuffdiff producing q-values > 1
Code:
Quote:
cuffdiff -p $NSLOTS -o cuffdiff_v110 Ensembl_59.gtf C_3732_hits.bam,C_6341_hits.bam,C_6356_hits.ba
m,C_6124_hits.bam,C_60PO_hits.bam HD_4430_hits.bam,HD_4412_hits.bam,HD_4242_hits.bam,HD_4189_hits
.bam,HD_3584_hits.bam
Versions
v091 v101 v102 v103 v110
OK
18112 33215 33230 33216 6765
Fail
3957 2524 2509 2523 1234
LOWDATA
NA 514 514 514 9000
NOTEST
29582 15408 15408 15408 34634
Total
51651 51661 51661 51661 51633
We wouldn’t expect such a big difference[1] between v110 and v101(v102,v103)especially in the numbers of NOTEST and OK. Cole Trapnell explained what does “Fail” mean [2], we would like to know what “NOTEST” means here.
All default options for the two versions appear to be the same; the command line options are exactly the same when we use different versions.
The cuffdiff v1.1.0 indicated.
“Cuffdiff now includes a more sophisticated check for sufficient sequencing depth prior to testingfor differences, which substantially improves the accuracy of differential expression analysis in loci with low to medium depth.”
Question on Cuffdiff’s q-value.
What is q-value here? Lots of evidence suggests q-value should be less than one [4], [5].
Reference:
(1) Inconsistency between cuffdiff 1.1.0 and cuffdiff 1.0.2
(2) Cole Trapnell on “Fail” test
(3) Cuffdiff producing q-values > 1
(4) Understanding q-values as a More Intuitive Alternative to p-values
(5) P-values, False Discovery Rate (FDR) and q-values What are p-values?
(6) Cuffdiff producing q-values > 1