Limits...
i-cisTarget 2015 update: generalized cis-regulatory enrichment analysis in human, mouse and fly.

Imrichová H, Hulselmans G, Atak ZK, Potier D, Aerts S - Nucleic Acids Res. (2015)

Bottom Line: As experimental data tracks we include transcription factor ChIP-seq data, histone modification ChIP-seq data and open chromatin data.The underlying processing method is based on a ranking-and-recovery procedure, allowing accurate determination of enrichment across heterogeneous datasets, while also discriminating direct from indirect target regions through a 'leading edge' analysis.Use of i-cisTarget is free and open to all, and there is no login requirement.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Computational Biology, Center for Human Genetics, University of Leuven, 3000 Leuven, Belgium.

Show MeSH

Related in: MedlinePlus

Ewing sarcoma case study. Various i-cisTarget analyses with different types of input, all related to EWS-FLI1 targets in Ewing sarcoma. (a) All input datasets are derived from Riggi et al. (35) and include FLI1 ChIP-seq peaks, top differentially less active peaks based on H3K27ac ChIP-seq upon EWS-FLI1 knockdown, differentially more open regions based on ATAC-seq after EWS-FLI1 activation and differentially downregulated genes after EWS-FLI1 knockdown (see ‘Materials and Methods’ section). (b) Input regions are automatically mapped to CRRs. (c). Each of the sets was analysed independently and reassuringly the expected motif EWS-FLI1 was ranked at the top, alongside regulatory tracks, mainly obtained on the SK-N-MC Ewing sarcoma cell line. Note that the rank of the motifs is represented by two values––the first one for the rank of the cluster of similar motifs, the second one is between brackets and represents the rank of the specific motif. (d) Distributions of AUC scores for a given input across all features in the selected databases (marked in purple in tables (c)) with an arrow indicating the top feature within that database. (e) The recovery curves for the top ranked features within the database, where the leading edge (LE) indicates the number of highly ranked target regions. (f) UCSC genome browser screenshot representing an example of one direct target region (red arrowhead) in the intron of gene APOH, which is included in the set of downregulated genes. This region was predicted as a target of EWS-FLI1 in i-cisTarget analyses, of the top less active H3K37ac peaks and the top 200 downregulated genes as well as FLI1 ChIP-seq peaks. The specific binding site is represented by a cluster of EWS-FLI1 motifs (green arrowhead), which was generated by i-cisTarget subsequent analysis, when the predicted target regions of EWS-FLI1 were scanned for CRMs of this factor. All these tracks are represented on the screenshot (from top to bottom): the CRRs, the predicted cluster of EWS-FLI1 motifs, RNA-seq peaks in SK-N-MC and A673 after shFLI1 (two purple tracks, published in (35)) and control (two orange tracks, published in (35)), H3K27ac peaks in SK-N-MC and MSC cell lines expressing EWS-FLI1 (green tracks, published in (35)), FLI1 ChIP-seq track in SK-N-MC and A673 cell lines (blue tracks, published in (35)) as well as DHS on SK-N-MC which was found as the top track within non-TF regulatory tracks (black track, from ENCODE database (15)).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4489282&req=5

Figure 2: Ewing sarcoma case study. Various i-cisTarget analyses with different types of input, all related to EWS-FLI1 targets in Ewing sarcoma. (a) All input datasets are derived from Riggi et al. (35) and include FLI1 ChIP-seq peaks, top differentially less active peaks based on H3K27ac ChIP-seq upon EWS-FLI1 knockdown, differentially more open regions based on ATAC-seq after EWS-FLI1 activation and differentially downregulated genes after EWS-FLI1 knockdown (see ‘Materials and Methods’ section). (b) Input regions are automatically mapped to CRRs. (c). Each of the sets was analysed independently and reassuringly the expected motif EWS-FLI1 was ranked at the top, alongside regulatory tracks, mainly obtained on the SK-N-MC Ewing sarcoma cell line. Note that the rank of the motifs is represented by two values––the first one for the rank of the cluster of similar motifs, the second one is between brackets and represents the rank of the specific motif. (d) Distributions of AUC scores for a given input across all features in the selected databases (marked in purple in tables (c)) with an arrow indicating the top feature within that database. (e) The recovery curves for the top ranked features within the database, where the leading edge (LE) indicates the number of highly ranked target regions. (f) UCSC genome browser screenshot representing an example of one direct target region (red arrowhead) in the intron of gene APOH, which is included in the set of downregulated genes. This region was predicted as a target of EWS-FLI1 in i-cisTarget analyses, of the top less active H3K37ac peaks and the top 200 downregulated genes as well as FLI1 ChIP-seq peaks. The specific binding site is represented by a cluster of EWS-FLI1 motifs (green arrowhead), which was generated by i-cisTarget subsequent analysis, when the predicted target regions of EWS-FLI1 were scanned for CRMs of this factor. All these tracks are represented on the screenshot (from top to bottom): the CRRs, the predicted cluster of EWS-FLI1 motifs, RNA-seq peaks in SK-N-MC and A673 after shFLI1 (two purple tracks, published in (35)) and control (two orange tracks, published in (35)), H3K27ac peaks in SK-N-MC and MSC cell lines expressing EWS-FLI1 (green tracks, published in (35)), FLI1 ChIP-seq track in SK-N-MC and A673 cell lines (blue tracks, published in (35)) as well as DHS on SK-N-MC which was found as the top track within non-TF regulatory tracks (black track, from ENCODE database (15)).

Mentions: As an example we use the ChIP-seq peaks of EWS-FLI1 as input. The authors ChIP'ped the C-terminal portion of FLI1 gene in two Ewing sarcoma cell lines (A673 and SK-N-MC) and defined the 1785 peaks that are present in both cell lines as the ‘core set of EWS-FLI1’ binding sites (provided in the Supplementary Table S1 of the corresponding paper). i-cisTarget analysis on this core set of binding sites identified the motif of the fusion product as the first motif (Figure 2c, i-cisTarget results on the website). Interestingly, ENCODE tracks of PolII ChIP-seq and DNAseI-seq on the Ewing sarcoma cell line SK-N-MC are highly ranked in the results. In addition, the remaining enriched features indicate a vast presence for ETS-family transcription factor motifs and this is indeed in agreement with the authors’ claim that the EWS-FLI1 oncogenic protein displaces ETS-family members from their native binding sites.


i-cisTarget 2015 update: generalized cis-regulatory enrichment analysis in human, mouse and fly.

Imrichová H, Hulselmans G, Atak ZK, Potier D, Aerts S - Nucleic Acids Res. (2015)

Ewing sarcoma case study. Various i-cisTarget analyses with different types of input, all related to EWS-FLI1 targets in Ewing sarcoma. (a) All input datasets are derived from Riggi et al. (35) and include FLI1 ChIP-seq peaks, top differentially less active peaks based on H3K27ac ChIP-seq upon EWS-FLI1 knockdown, differentially more open regions based on ATAC-seq after EWS-FLI1 activation and differentially downregulated genes after EWS-FLI1 knockdown (see ‘Materials and Methods’ section). (b) Input regions are automatically mapped to CRRs. (c). Each of the sets was analysed independently and reassuringly the expected motif EWS-FLI1 was ranked at the top, alongside regulatory tracks, mainly obtained on the SK-N-MC Ewing sarcoma cell line. Note that the rank of the motifs is represented by two values––the first one for the rank of the cluster of similar motifs, the second one is between brackets and represents the rank of the specific motif. (d) Distributions of AUC scores for a given input across all features in the selected databases (marked in purple in tables (c)) with an arrow indicating the top feature within that database. (e) The recovery curves for the top ranked features within the database, where the leading edge (LE) indicates the number of highly ranked target regions. (f) UCSC genome browser screenshot representing an example of one direct target region (red arrowhead) in the intron of gene APOH, which is included in the set of downregulated genes. This region was predicted as a target of EWS-FLI1 in i-cisTarget analyses, of the top less active H3K37ac peaks and the top 200 downregulated genes as well as FLI1 ChIP-seq peaks. The specific binding site is represented by a cluster of EWS-FLI1 motifs (green arrowhead), which was generated by i-cisTarget subsequent analysis, when the predicted target regions of EWS-FLI1 were scanned for CRMs of this factor. All these tracks are represented on the screenshot (from top to bottom): the CRRs, the predicted cluster of EWS-FLI1 motifs, RNA-seq peaks in SK-N-MC and A673 after shFLI1 (two purple tracks, published in (35)) and control (two orange tracks, published in (35)), H3K27ac peaks in SK-N-MC and MSC cell lines expressing EWS-FLI1 (green tracks, published in (35)), FLI1 ChIP-seq track in SK-N-MC and A673 cell lines (blue tracks, published in (35)) as well as DHS on SK-N-MC which was found as the top track within non-TF regulatory tracks (black track, from ENCODE database (15)).
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4489282&req=5

Figure 2: Ewing sarcoma case study. Various i-cisTarget analyses with different types of input, all related to EWS-FLI1 targets in Ewing sarcoma. (a) All input datasets are derived from Riggi et al. (35) and include FLI1 ChIP-seq peaks, top differentially less active peaks based on H3K27ac ChIP-seq upon EWS-FLI1 knockdown, differentially more open regions based on ATAC-seq after EWS-FLI1 activation and differentially downregulated genes after EWS-FLI1 knockdown (see ‘Materials and Methods’ section). (b) Input regions are automatically mapped to CRRs. (c). Each of the sets was analysed independently and reassuringly the expected motif EWS-FLI1 was ranked at the top, alongside regulatory tracks, mainly obtained on the SK-N-MC Ewing sarcoma cell line. Note that the rank of the motifs is represented by two values––the first one for the rank of the cluster of similar motifs, the second one is between brackets and represents the rank of the specific motif. (d) Distributions of AUC scores for a given input across all features in the selected databases (marked in purple in tables (c)) with an arrow indicating the top feature within that database. (e) The recovery curves for the top ranked features within the database, where the leading edge (LE) indicates the number of highly ranked target regions. (f) UCSC genome browser screenshot representing an example of one direct target region (red arrowhead) in the intron of gene APOH, which is included in the set of downregulated genes. This region was predicted as a target of EWS-FLI1 in i-cisTarget analyses, of the top less active H3K37ac peaks and the top 200 downregulated genes as well as FLI1 ChIP-seq peaks. The specific binding site is represented by a cluster of EWS-FLI1 motifs (green arrowhead), which was generated by i-cisTarget subsequent analysis, when the predicted target regions of EWS-FLI1 were scanned for CRMs of this factor. All these tracks are represented on the screenshot (from top to bottom): the CRRs, the predicted cluster of EWS-FLI1 motifs, RNA-seq peaks in SK-N-MC and A673 after shFLI1 (two purple tracks, published in (35)) and control (two orange tracks, published in (35)), H3K27ac peaks in SK-N-MC and MSC cell lines expressing EWS-FLI1 (green tracks, published in (35)), FLI1 ChIP-seq track in SK-N-MC and A673 cell lines (blue tracks, published in (35)) as well as DHS on SK-N-MC which was found as the top track within non-TF regulatory tracks (black track, from ENCODE database (15)).
Mentions: As an example we use the ChIP-seq peaks of EWS-FLI1 as input. The authors ChIP'ped the C-terminal portion of FLI1 gene in two Ewing sarcoma cell lines (A673 and SK-N-MC) and defined the 1785 peaks that are present in both cell lines as the ‘core set of EWS-FLI1’ binding sites (provided in the Supplementary Table S1 of the corresponding paper). i-cisTarget analysis on this core set of binding sites identified the motif of the fusion product as the first motif (Figure 2c, i-cisTarget results on the website). Interestingly, ENCODE tracks of PolII ChIP-seq and DNAseI-seq on the Ewing sarcoma cell line SK-N-MC are highly ranked in the results. In addition, the remaining enriched features indicate a vast presence for ETS-family transcription factor motifs and this is indeed in agreement with the authors’ claim that the EWS-FLI1 oncogenic protein displaces ETS-family members from their native binding sites.

Bottom Line: As experimental data tracks we include transcription factor ChIP-seq data, histone modification ChIP-seq data and open chromatin data.The underlying processing method is based on a ranking-and-recovery procedure, allowing accurate determination of enrichment across heterogeneous datasets, while also discriminating direct from indirect target regions through a 'leading edge' analysis.Use of i-cisTarget is free and open to all, and there is no login requirement.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Computational Biology, Center for Human Genetics, University of Leuven, 3000 Leuven, Belgium.

Show MeSH
Related in: MedlinePlus