Limits...
Chromatin segmentation based on a probabilistic model for read counts explains a large portion of the epigenome.

Mammana A, Chung HR - Genome Biol. (2015)

Bottom Line: Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is an increasingly common experimental approach to generate genome-wide maps of histone modifications and to dissect the complexity of the epigenome.Here, we propose EpiCSeg: a novel algorithm that combines several histone modification maps for the segmentation and characterization of cell-type specific epigenomic landscapes.By using an accurate probabilistic model for the read counts, EpiCSeg provides a useful annotation for a considerably larger portion of the genome, shows a stronger association with validation data, and yields more consistent predictions across replicate experiments when compared to existing methods.The software is available at http://github.com/lamortenera/epicseg.

View Article: PubMed Central - PubMed

Affiliation: Otto-Warburg-Laboratory, Epigenomics, Max Planck Institute for Molecular Genetics, D-14195, Berlin, Germany. mammana@molgen.mpg.de.

ABSTRACT
Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is an increasingly common experimental approach to generate genome-wide maps of histone modifications and to dissect the complexity of the epigenome. Here, we propose EpiCSeg: a novel algorithm that combines several histone modification maps for the segmentation and characterization of cell-type specific epigenomic landscapes. By using an accurate probabilistic model for the read counts, EpiCSeg provides a useful annotation for a considerably larger portion of the genome, shows a stronger association with validation data, and yields more consistent predictions across replicate experiments when compared to existing methods.The software is available at http://github.com/lamortenera/epicseg.

No MeSH data available.


Related in: MedlinePlus

Chromatin states-based prediction of transcription levels. The value to be predicted is the log-transformed RNA-seq coverage per bin and the predictor is the chromatin state per bin. The R2 values were computed using standard linear regression
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4514447&req=5

Fig3: Chromatin states-based prediction of transcription levels. The value to be predicted is the log-transformed RNA-seq coverage per bin and the predictor is the chromatin state per bin. The R2 values were computed using standard linear regression

Mentions: Next, we measured how well chromatin states can predict gene expression. For that purpose we used a cell-type specific RNA-seq experiment for each dataset. As a measure of gene expression levels we used the logarithm of the average RNA-seq coverage per bin (adding a pseudo-count of 1) and as a measure for predictive power we computed the R2 resulting from standard linear regression with a categorical predictor (the chromatin states). FigureĀ 3 shows that EpiCSeg and ChromHMM have a similar predictive power, but the former tends to perform better. The low R2 values observed in the IMR90 and H1 datasets might suggest that in datasets with many ChIP-seq tracks the segmentation algorithms are less influenced by transcription-associated histone marks (for example, H3K36me3).Fig. 3


Chromatin segmentation based on a probabilistic model for read counts explains a large portion of the epigenome.

Mammana A, Chung HR - Genome Biol. (2015)

Chromatin states-based prediction of transcription levels. The value to be predicted is the log-transformed RNA-seq coverage per bin and the predictor is the chromatin state per bin. The R2 values were computed using standard linear regression
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4514447&req=5

Fig3: Chromatin states-based prediction of transcription levels. The value to be predicted is the log-transformed RNA-seq coverage per bin and the predictor is the chromatin state per bin. The R2 values were computed using standard linear regression
Mentions: Next, we measured how well chromatin states can predict gene expression. For that purpose we used a cell-type specific RNA-seq experiment for each dataset. As a measure of gene expression levels we used the logarithm of the average RNA-seq coverage per bin (adding a pseudo-count of 1) and as a measure for predictive power we computed the R2 resulting from standard linear regression with a categorical predictor (the chromatin states). FigureĀ 3 shows that EpiCSeg and ChromHMM have a similar predictive power, but the former tends to perform better. The low R2 values observed in the IMR90 and H1 datasets might suggest that in datasets with many ChIP-seq tracks the segmentation algorithms are less influenced by transcription-associated histone marks (for example, H3K36me3).Fig. 3

Bottom Line: Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is an increasingly common experimental approach to generate genome-wide maps of histone modifications and to dissect the complexity of the epigenome.Here, we propose EpiCSeg: a novel algorithm that combines several histone modification maps for the segmentation and characterization of cell-type specific epigenomic landscapes.By using an accurate probabilistic model for the read counts, EpiCSeg provides a useful annotation for a considerably larger portion of the genome, shows a stronger association with validation data, and yields more consistent predictions across replicate experiments when compared to existing methods.The software is available at http://github.com/lamortenera/epicseg.

View Article: PubMed Central - PubMed

Affiliation: Otto-Warburg-Laboratory, Epigenomics, Max Planck Institute for Molecular Genetics, D-14195, Berlin, Germany. mammana@molgen.mpg.de.

ABSTRACT
Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is an increasingly common experimental approach to generate genome-wide maps of histone modifications and to dissect the complexity of the epigenome. Here, we propose EpiCSeg: a novel algorithm that combines several histone modification maps for the segmentation and characterization of cell-type specific epigenomic landscapes. By using an accurate probabilistic model for the read counts, EpiCSeg provides a useful annotation for a considerably larger portion of the genome, shows a stronger association with validation data, and yields more consistent predictions across replicate experiments when compared to existing methods.The software is available at http://github.com/lamortenera/epicseg.

No MeSH data available.


Related in: MedlinePlus