(a) Genomic tracks display ChIP-seq data (I, Illumina; H, Helicos). Shown are enrichment profiles for CTCF (black), H3K4me3 (green), H3K27me3 (red) and H3K36me3 (blue) across a 300kb region in mouse ES cells. The UCSC known genes track is shown below the profiles. (b) Quantitative comparison of histone modifications from ChIP data sequenced by Illumina or Helicos. Scatter plots show signals for H3K4me3 (1kb bins), H3K27me3 (5kb bins) and H3K36me3 (5kb bins) across the genome. Indicated on each plot is the Pearson correlation coefficient, ρ. (c) Quantitative comparison of CTCF ChIP-Seq data. Venn diagram indicates overlap between top 20,000 genomic locations bound by CTCF as determined by Illumina (red) or Helicos (blue). The comparisons confirm strong concordance for data collected on the respective platforms.

Normalized read density obtained from un-enriched control ‘whole cell extract’ (WCE) samples analyzed by Illumina (blue) or Helicos (black) is plotted as a function of GC percentage for 100bp windows. A theoretical ‘expected’ distribution obtained by computational simulation is shown in red (Random). The Illumina data modestly over-represent reads from regions with a GC-content of ~40–65%, while the Helicos data show a relatively even distribution across 20–80% GC-content.

(a) Genomic view displays H3K4me3 ChIP-seq data obtained with either Illumina (I), standard Helicos (H) or small sample size Helicos methodology (H 150pg and H 50pg). Shown are enrichment signals across a 440kb region in mouse ES cells. The UCSC known genes track appears below the profiles. (b–c) ChIP-Seq data derived from 50 picograms of sample compared against data derived from 3 nanogram or 150 picogram samples. Scatter plots show H3K4me3 signal for 1 kb bins across the genome. Pearson correlation, ρ, is indicated for each comparison. (d) Genomic view displays ChIP-seq data obtained with either Illumina (I), standard Helicos (H) or small sample size Helicos methodology (H 200pg) for a 300kb regions in mouse ES cells. These datasets were not truncated and thus have different read numbers. (e) Quantitative comparison of CTCF ChIP-Seq data derived by sequencing either 200 picograms or 3 nanograms of ChIP DNA by Helicos. Venn diagram indicates overlap between the top 20,000 enriched genomic locations in each dataset.