MethBat v0.17.0

Changes

• Added new methbat report sub-command for analyzing pre-defined regions with known expected methylation patterns (e.g., imprinting regions). The command compares observed methylation patterns against expected patterns, identifying regions with anomalous methylation states and generating quality control warnings. See report guide for details on usage.
• Added example report region files for GRCh38 imprinting regions in data/report_regions
• Added an output header to most output CSV/TSV files that includes the MethBat version, command, and the datetime the command was run. These header lines are prefixed with the '#' character.

Fixed

• The profile output TSV files have been modified such that their column headers do not have the '#' prefix

---

This release was automatically updated from tag v0.17.0.

MethBat v0.16.1

Fixed

• Fixed an issue where compressed bed files were not recognized by methbat signature mode

---

This release was automatically updated from tag v0.16.1.

MethBat v0.16.0

Changes

• Added new methbat deconvolve sub-command for cell type deconvolution from bulk methylation data. See deconvolution guide for details on usage.
• Added two example atlas files derived from public atlases: data/cell_atlas

---

This release was automatically updated from tag v0.16.0.

MethBat v0.15.0

Changes

• Breaking: Added five new columns to the output of profile:
  • cpg_label - A pass-through of the optional cpg_label field in the provided regions or background file
  • mean_combined_methyl_delta - This is the raw delta value that is used to generate mean_combined_methyl_zscore. This will only be populated if background statistics are provided.
  • median_total_coverage, median_hap1_coverage, and median_hap2_coverage - These metrics are calculated by extracting the coverage values from the pb-CpG-tools inputs and calculating the median of each. For hap1 and hap2, only the CpGs with phased information will be used.
• Breaking: Added the cpg_label column to the build output, this is a pass-through of the label from the provided profiles

---

This release was automatically updated from tag v0.15.0.

MethBat v0.14.2

Fixed

• Fixed an issue where CpG segments with fewer CpGs than the --min-cpgs parameter could get created. This primarily occurred with groups of isolated CpGs, and affected segment, joint-segment, and signature modes. These small isolated segments are now filtered after the segmentation step, and therefor removed from all outputs.

MethBat v0.14.1

Fixed

• Improved error handling for small population sizes when calling signature
• Added earlier error checks with better messaging for signature

MethBat v0.14.0

Changes

• Adds a new sub-command methbat joint-segment: This sub-command loads all methylation values for a cohort and generates an average methylation for the collection. This average is then segmented into combined methylation and allele-specific methylation segments.
• Breaking change - The CLI verbosity flag (-v) has been moved into a shared location. Commands that previously used it (e.g. methbat segment -v ...) will need to update the position to before the sub-command (e.g. methbat -v segment ...).

MethBat v0.13.3

Changes

• Added support for parsing pb-CpG-tools files that have been compressed (i.e., {prefix}.combined.bed.gz) to save storage

MethBat v0.13.2

Fixed

• Modified the checker for profile and segment modes to allow for missing hap1 and hap2 BED files. This is most common when pb-CpG-tools is run on unphased datasets. A warning will be produced in the log if these files are missing, and haplotype-specific data will not be generated.

MethBat v0.13.1

Changes

• Adds two haplotype-specific tracks for methbat segment:
  • {OUT_PREFIX}.hap1.bedgraph - Contains the underlying methylation values for haplotype 1 for each segmented region
  • {OUT_PREFIX}.hap2.bedgraph - Same as above, but for haplotype 2

Fixed

• Fixed an issue where the BEDGRAPH outputs from methbat segment were reporting median values instead of the indicated mean values. This change did not alter the segmentation itself. However, since the value associated with each segment changed, the reported regions in {OUT_PREFIX}.meth_regions.bed may be altered to reflect the correct, mean values.