Erratum to: Chronic cocaine-regulated epigenomic changes in mouse nucleus accumbens

During the typesetting of the final version of the article [1] Tables S2, S8 and S11 were duplicated and Tables S1, S7 and S9 are missing. We apologize for the mistake that led to the loss of some of the additional files.

During the typesetting of the final version of the article [1] Tables S2, S8 and S11 were duplicated and Tables S1, S7 and S9 are missing. We apologize for the mistake that led to the loss of some of the additional files.
In the final version of the article: Additional file 2: Table S1 is wrongly replaced as Table S2.
Additional file 17: Table S7 is wrongly replaced as Table S8.
Additional file 22: Table S9 is wrongly replaced as Table S11 Please see missing Additional files 2, 17 and 22 below. All Additional files were correct in the provisional version of the article. Below please find the correct full list of Additional Files associated with this article.
Additional file 2: Table S1. RNA-seq quality control metrics for acute and chronic data.
Additional file 3: Table S2. Differential RNA-seq lists. Differential gene lists from repeated and acute cocaine RNAseq experiments; differential splicing lists from repeated cocaine RNA-seq experiments. Cuffdiff was used to perform differential analysis for various transcriptomic events.
Additional file 5: Figure S3. RNAseq nanostring validation. Nanostring validation of cocaine-induced changes in RNA expression in NAc. A separate cohort of animals was used to validate RNA-seq results. Normalized Nanostring read counts are shown on the y-axis. All genes display the same direction of change with significance as seen with RNA-seq. Error bars are mean ± standard error of the mean derived from 14 cocaine treated and 14 saline treated samples. *P < 0.05, **P < 0.01.
Additional file 6: Table S3. GO term enrichment of genes that have altered splicing. The altered splicing group combines the genes that contain alternative promoter usage and/or alternative splicing. DAVID is used to perform GO analysis. Only three GO categories are used: biological process, cellular component, and molecular function.
Additional file 9: Figure S4. Global enrichment plots and numbers of differential events. Each panel includes five sub-figures for the enrichment, using data pooled from the three biological replicates, of an epigenomic mark at TSSs, gene bodies, transcriptional end sites, and cocaine up-regulated sites and down-regulated sites. Y-axes represent the normalized coverage (RPM) that is averaged across all genomic regions. Additional file 10: Table S5. Differential sites for the seven epigenomic marks. diffReps is used to identify differential sites for each of the seven epigenomic marks. A FDR cutoff of <10 % was used to choose the sites that are significant.
Additional file 11: Figure S5 Additional file 12: Figure S6. Heatmap showing the enrichment of the top 30 and custom pathways among the seven marks. The darkness of each grid represents the statistical significance of enrichment.
Additional file 13: Table S6. Enrichment analysis of ChIP-seq differential sites. After the differential sites are mapped to promoter or gene body, the genes that contain the differential sites are uploaded to IPA for enrichment analysis. The enriched canonical pathways or customized gene lists are extracted. Each value represents -log10(P-value) of enrichment. Co-occurrence score (Additional file 7) is used to rank the pathways in descending order.
Additional file 14: Figure S7. Differential sites to exon center distance density plots (related to Figure 3). The distance between each differential site and the closest exon center was calculated. The exons were further classified into three categories: promoter, internal, and polyA. The density for the distance within a 10 kb window of the exon center of each type was calculated. Each panel represents an epigenomic mark. Additional file 16: Figure S9. Construction of cocaine-induced chromatin signatures. All chromatin signatures are put into a signature matrix with each row being a transcript and each column being the log fold change of each mark at each genomic region. K-means clustering was performed on the signature matrix to group transcripts into signature clusters that share common chromatin modification patterns. The regions that show significant chromatin changes were extracted to perform motif analysis to identify potential splicing and transcription factors.
Additional file 17: Table S7. Genome-wide association between chromatin modification and transcriptional change. Based on the chromatin modification at each genomic region, transcripts are separated into up-regulated, down-regulated, and non-significant (Additional file 7). The chromatin-up and -down transcripts are correlated with transcripts that show expression change using Fisher's exact test. This generates four combinations ('s' = chromatin modification, 'e' = expression change): s.up.e.up; s.up.e.down; s.down.e.up; s.down.e.down. The P-values were adjusted using the BH [78] method and a FDR cutoff of <10 % was used to select markregion combinations. The analysis was first done with the enhancer regions included and then repeated with the enhancers removed.
Additional file 18: Figure S11. Chromatin modification heatmap for 29 signature clusters. A merged heatmap for all 29 signature clusters with transcripts as rows and mark-region combinations as columns. The color key indicates log2 fold changes. Different clusters are labeled by different colors.
Additional file 19: Table S8. Enriched functional terms and canonical pathways among the 29 signature clusters. IPA was used to find the enriched biological functions and canonical pathways among the signature clusters. The co-occurrence score was then used to rank the enriched terms in descending order.
Additional file 20: Motif intermediate results. This zip file contains the textual outputs from motif analysis. The motifs found by MEME [42] were first combined using the Bayesian motif clustering [79] method and then matched with known motifs. Further explanations are provided in the enclosed README file.
Additional file 21: Figure S10. Quantitative ChIP validation of cocaine-induced changes in H3K4me3 in NAc. A separate cohort of animals was used to validate ChIP-seq data. All genomic sites tested display the same direction of change with significance as seen