Skip to main content

Table 1 Alternative splicing events

From: SpliceGrapher: detecting patterns of alternative splicing from RNA-Seq data in the context of gene models and EST data

 

AS

AS events

 

genes

IR

ES

Alt. 5'

Alt. 3'

Total

A. thaliana models

4,029

1,987 (33%)

550 (9%)

1,256 (21%)

2,145 (36%)

5,938

   SpliceGrapher

      

No ESTs

4,901

2,248 (30%)

714 (10%)

1,560 (21%)

2,866 (39%)

7,388

Novel

885

308 (21%)

164 (11%)

304 (20%)

721 (48%)

1,497

With ESTs

6,162

3,658 (33%)

994 (9%)

2,335 (21%)

4,128 (37%)

9,916

Novel

2,154

1,779 (34%)

444 (8%)

1,079 (20%)

1,983 (38%)

5,285

   Cufflinks

      

No gene models

1,263

449 (32%)

383 (28%)

237 (17%)

319 (23%)

1,388

Novel

699

429 (32%)

380 (28%)

232 (17%)

304 (23%)

1,345

With gene models

6,056

4,029 (39%)

2,857 (27%)

1,427 (14%)

2,106 (20%)

10,419

Novel

2,319

2,232 (38%)

2,550 (43%)

552 (9%)

594 (10%)

5,928

   TAU

      

No gene models

2,777

893 (17%)

475 (9%)

1,481 (27%)

2,555 (47%)

5,404

Novel

1,591

811 (16%)

460 (9%)

1,431 (28%)

2,351 (47%)

5,053

With gene models

10,458

94,571 (85%)

598 (1%)

5,972 (5%)

9,820 (9%)

110,961

Novel

8,364

94,124 (86%)

476 (0%)

5,697 (5%)

9,219 (8%)

109,516

V. vinifera models

0

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0

   SpliceGrapher

2,039

347 (13%)

830 (31%)

640 (24%)

838 (32%)

2,655

   TAU

      

No gene models

3,099

531 (10%)

684 (13%)

1,321 (25%)

2,743 (52%)

5,279

With gene models

15,874

135,585 (72%)

4,938 (3%)

23,615 (13%)

24,406 (13%)

188,544

   Cufflinks

      

No gene models

1,057

324 (24%)

519 (39%)

140 (11%)

349 (26%)

1,332

With gene models

4,263

4,120 (34%)

3,148 (26%)

2,165 (18%)

2,818 (23%)

12,251

  1. The number of AS events detected by SpliceGrapher, Cufflinks, and TAU compared with events inferred from the TAIR9 annotations. We track the following AS event types: intron retention (IR), exon skipping (ES), alternative 5' sites (Alt. 5') and alternative 3' sites (Alt. 3'). SpliceGrapher uses the TAIR9 gene models as a baseline, so it includes all of the same AS events along with additional events inferred from RNA-Seq data. Without gene models, nearly all TAU and Cufflinks predictions are novel AS events. With gene models, more than half of Cufflinks predictions reproduce AS events from the gene models. TAU uses known splice sites to predict all possible exons in a gene, generating vast numbers of novel exons and novel IR events.