Limits...
Annotation of genomics data using bidirectional hidden Markov models unveils variations in Pol II transcription cycle.

Zacher B, Lidschreiber M, Cramer P, Gagneur J, Tresch A - Mol. Syst. Biol. (2014)

Bottom Line: To overcome these limitations, we introduce bidirectional HMMs which infer directed genomic states from occupancy profiles de novo.Application to RNA polymerase II-associated factors in yeast and chromatin modifications in human T cells recovers the majority of transcribed loci, reveals gene-specific variations in the yeast transcription cycle and indicates the existence of directed chromatin state patterns at transcribed, but not at repressed, regions in the human genome.We anticipate bidirectional HMMs to significantly improve the analyses of genome-associated directed processes.

View Article: PubMed Central - PubMed

Affiliation: Gene Center and Department of Biochemistry, Center for Integrated Protein Science CIPSM, Ludwig-Maximilians-Universität München, Munich, Germany Institute for Genetics, University of Cologne, Cologne, Germany.

No MeSH data available.


Related in: MedlinePlus

Genomic state annotation predicts bidirectional promoters and (novel) transcripts.The genomic state annotation (viterbi path) was searched with regular expressions (RegEx) defining bidirectional promoters (right) and transcripts (bottom).Nucleosome binding patterns centered at 1,076 identified bidirectional promoters found with the RegEx. Each line in the heatmap corresponds to one pair of transcripts. Binding signal is color-coded (right).A novel SUT (stable unannotated transcript, a stable non-coding RNA, gray area) is identified on the − strand by the bdHMM. The locus shows detectable expression but was too low for the criteria used by Xu et al (2009).Estimated cumulative probability of TSS and pA site predictions shows higher accuracy of bdHMM in recovering TSSs. pA site prediction has similar accuracy for both models.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4300491&req=5

fig03: Genomic state annotation predicts bidirectional promoters and (novel) transcripts.The genomic state annotation (viterbi path) was searched with regular expressions (RegEx) defining bidirectional promoters (right) and transcripts (bottom).Nucleosome binding patterns centered at 1,076 identified bidirectional promoters found with the RegEx. Each line in the heatmap corresponds to one pair of transcripts. Binding signal is color-coded (right).A novel SUT (stable unannotated transcript, a stable non-coding RNA, gray area) is identified on the − strand by the bdHMM. The locus shows detectable expression but was too low for the criteria used by Xu et al (2009).Estimated cumulative probability of TSS and pA site predictions shows higher accuracy of bdHMM in recovering TSSs. pA site prediction has similar accuracy for both models.

Mentions: In order to re-annotate transcription throughout the yeast genome and compare the performance of bdHMM and HMM, we applied a regular expression (RegEx) approach (Fig3A), to predict transcribed units as continuous stretches of directed transcribed states with a minimal length of 80 bp on both strands from the bdHMM and HMM annotation. Matching predicted transcript boundaries to previously published ones (Xu et al, 2009), 4,186 (82%) of all annotated protein-coding transcripts were recovered from the bdHMM predictions, 11% more than the HMM predicts using the same criteria (3,639 transcripts) (best reciprocal hits, Materials and Methods). Moreover, the predicted transcription start sites (TSS) were consistently closer to the annotated ones (Fig3D). In particular, 60% of the predicted TSSs by the bdHMM were within 50 bp, whereas the best 60% of the HMM TSS predictions were within 100 bp of the published ones. Accuracy of pA site prediction was lower, but comparable between bdHMM and HMM, where approximately 60% of the predicted pA sites were within 100 bp of the annotated ones for both methods. Moreover, 32 novel transcripts were predicted from the bdHMM annotation (four overlapping a coding region, 28 non-coding, Fig3C, Materials and Methods), which is of particular significance because the S. cerevisiae transcriptome has been thoroughly studied and annotated.


Annotation of genomics data using bidirectional hidden Markov models unveils variations in Pol II transcription cycle.

Zacher B, Lidschreiber M, Cramer P, Gagneur J, Tresch A - Mol. Syst. Biol. (2014)

Genomic state annotation predicts bidirectional promoters and (novel) transcripts.The genomic state annotation (viterbi path) was searched with regular expressions (RegEx) defining bidirectional promoters (right) and transcripts (bottom).Nucleosome binding patterns centered at 1,076 identified bidirectional promoters found with the RegEx. Each line in the heatmap corresponds to one pair of transcripts. Binding signal is color-coded (right).A novel SUT (stable unannotated transcript, a stable non-coding RNA, gray area) is identified on the − strand by the bdHMM. The locus shows detectable expression but was too low for the criteria used by Xu et al (2009).Estimated cumulative probability of TSS and pA site predictions shows higher accuracy of bdHMM in recovering TSSs. pA site prediction has similar accuracy for both models.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4300491&req=5

fig03: Genomic state annotation predicts bidirectional promoters and (novel) transcripts.The genomic state annotation (viterbi path) was searched with regular expressions (RegEx) defining bidirectional promoters (right) and transcripts (bottom).Nucleosome binding patterns centered at 1,076 identified bidirectional promoters found with the RegEx. Each line in the heatmap corresponds to one pair of transcripts. Binding signal is color-coded (right).A novel SUT (stable unannotated transcript, a stable non-coding RNA, gray area) is identified on the − strand by the bdHMM. The locus shows detectable expression but was too low for the criteria used by Xu et al (2009).Estimated cumulative probability of TSS and pA site predictions shows higher accuracy of bdHMM in recovering TSSs. pA site prediction has similar accuracy for both models.
Mentions: In order to re-annotate transcription throughout the yeast genome and compare the performance of bdHMM and HMM, we applied a regular expression (RegEx) approach (Fig3A), to predict transcribed units as continuous stretches of directed transcribed states with a minimal length of 80 bp on both strands from the bdHMM and HMM annotation. Matching predicted transcript boundaries to previously published ones (Xu et al, 2009), 4,186 (82%) of all annotated protein-coding transcripts were recovered from the bdHMM predictions, 11% more than the HMM predicts using the same criteria (3,639 transcripts) (best reciprocal hits, Materials and Methods). Moreover, the predicted transcription start sites (TSS) were consistently closer to the annotated ones (Fig3D). In particular, 60% of the predicted TSSs by the bdHMM were within 50 bp, whereas the best 60% of the HMM TSS predictions were within 100 bp of the published ones. Accuracy of pA site prediction was lower, but comparable between bdHMM and HMM, where approximately 60% of the predicted pA sites were within 100 bp of the annotated ones for both methods. Moreover, 32 novel transcripts were predicted from the bdHMM annotation (four overlapping a coding region, 28 non-coding, Fig3C, Materials and Methods), which is of particular significance because the S. cerevisiae transcriptome has been thoroughly studied and annotated.

Bottom Line: To overcome these limitations, we introduce bidirectional HMMs which infer directed genomic states from occupancy profiles de novo.Application to RNA polymerase II-associated factors in yeast and chromatin modifications in human T cells recovers the majority of transcribed loci, reveals gene-specific variations in the yeast transcription cycle and indicates the existence of directed chromatin state patterns at transcribed, but not at repressed, regions in the human genome.We anticipate bidirectional HMMs to significantly improve the analyses of genome-associated directed processes.

View Article: PubMed Central - PubMed

Affiliation: Gene Center and Department of Biochemistry, Center for Integrated Protein Science CIPSM, Ludwig-Maximilians-Universität München, Munich, Germany Institute for Genetics, University of Cologne, Cologne, Germany.

No MeSH data available.


Related in: MedlinePlus