gene and isoform expression level estimation with Cufflinks

[Jane M]

Hello,

I use Cufflinks to estimate gene but particularly isoform expression level estimation.

I have a GFF file:

DS571145	alternativeSplicer	gene	3484	4143	.	-	.	ID=EHI_151170;Name=hypothetical protein;
DS571145	alternativeSplicer	mRNA	3484	4143	.	-	.	ID=EHI_151170.ref;Name=EHI_151170.ref;Parent=EHI_151170;completeORF=yes
DS571145	alternativeSplicer	exon	3484	4143	.	-	.	ID=exon_EHI_151170.ref-1;Name=exon;Parent=EHI_151170.ref;
DS571145	alternativeSplicer	mRNA	3484	4143	.	-	.	ID=EHI_151170.alt1;Name=EHI_151170.alt1;Parent=EHI_151170;completeORF=no
DS571145	alternativeSplicer	exon	3484	3943	.	-	.	ID=exon_EHI_151170.alt1-1;Name=exon;Parent=EHI_151170.alt1;
DS571145	alternativeSplicer	exon	3997	4143	.	-	.	ID=exon_EHI_151170.alt1-2;Name=exon;Parent=EHI_151170.alt1;
DS571145	alternativeSplicer	gene	4256	5944	.	+	.	ID=EHI_151180;Name=calcineurin catalytic subunit A putative;
DS571145	alternativeSplicer	mRNA	4256	5944	.	+	.	ID=EHI_151180.ref;Name=EHI_151180.ref;Parent=EHI_151180;completeORF=yes
DS571145	alternativeSplicer	exon	4256	5944	.	+	.	ID=exon_EHI_151180.ref-1;Name=exon;Parent=EHI_151180.ref;
DS571145	alternativeSplicer	gene	9060	10194	.	-	.	ID=EHI_151210.1;Name=TBC domain containing protein;
DS571145	alternativeSplicer	mRNA	9060	10194	.	-	.	ID=EHI_151210.1.ref;Name=EHI_151210.1.ref;Parent=EHI_151210.1;completeORF=yes
DS571145	alternativeSplicer	exon	9060	9961	.	-	.	ID=exon_EHI_151210.1.ref-1;Name=exon;Parent=EHI_151210.1.ref;
DS571145	alternativeSplicer	exon	10053	10194	.	-	.	ID=exon_EHI_151210.1.ref-2;Name=exon;Parent=EHI_151210.1.ref;

I turned it into a GTF file and I used this file to run Cufflinks:

DS571145	alternativeSplicer	exon	3484	4143	.	-	.	gene_id "EHI_151170"; transcript_id "exon_EHI_151170.ref-1";
DS571145	alternativeSplicer	exon	3484	3943	.	-	.	gene_id "EHI_151170"; transcript_id "exon_EHI_151170.alt1-1";
DS571145	alternativeSplicer	exon	3997	4143	.	-	.	gene_id "EHI_151170"; transcript_id "exon_EHI_151170.alt1-2";
DS571145	alternativeSplicer	exon	4256	5944	.	+	.	gene_id "EHI_151180"; transcript_id "exon_EHI_151180.ref-1";
DS571145	alternativeSplicer	exon	9060	9961	.	-	.	gene_id "EHI_151210.1"; transcript_id "exon_EHI_151210.1.ref-1";
DS571145	alternativeSplicer	exon	10053	10194	.	-	.	gene_id "EHI_151210.1"; transcript_id "exon_EHI_151210.1.ref-2";

First, I'm not sure, should I keep the mRNA/gene lines?
This file contains 9864 lines, so 9864 exons.

As a result, I get the outputs of Cufflinks (genes.expr, transcripts.expr and transcripts.gtf) for each lane. Here is an example of one lane:

- gene.expr:

EHI_151170	7186	DS571145	3483	4143	27.9499	15.5065	40.3932	OK
EHI_151180	7187	DS571145	4255	5944	18.3523	9.7844	26.9202	OK
EHI_151210.1	7190	DS571145	9059	9961	33.9837	22.3246	45.6428	OK
EHI_151210.1	7191	DS571145	10052	10194	43.9839	30.7198	57.248	OK

First problem: some genes appear several times, because exon and gene are mixed (but not all!)...

- transcripts.expr:

exon_EHI_151170.alt1-1	7186	DS571145	3483	3943	0.267399	0.00180684	0.00108507	0	2.90168	0.112715	460	361	OK
exon_EHI_151170.ref-1	7186	DS571145	3483	4143	147.993	1	0.933245	122.898	173.088	62.3827	660	561	OK
exon_EHI_151170.alt1-2	7186	DS571145	3996	4143	121.712	0.822414	0.0656696	49.073	194.35	51.3044	147	48	OK
exon_EHI_151180.ref-1	7187	DS571145	4255	5944	176.861	1	1	150.263	203.459	78.6164	1689	1590	OK
exon_EHI_151210.1.ref-1	7190	DS571145	9059	9961	327.5	1	1	291.306	363.694	147.572	902	803	OK
exon_EHI_151210.1.ref-2	7191	DS571145	10052	10194	423.872	1	1	382.696	465.049	102.326	142	43	OK

My main problem is here, in this "transcripts.expr" file: the result is the exon expression level estimation, instead of isoform expression level estimation.
Moreover, I should find the same number of exons as in my GFT file, but this is not the case! 583 exons are lost!


Has someone managed to use Cufflinks to compute gene and isoform expression level estimation? Which GTF file did you use?
Any help would be appreciated, I'm confused ...

[Jane M]

The problem came from my GTF file (the transcript id).
Finally, Cufflinks is working well with the GFF file ...