StarPhase v1.0.0

Changes

• Source code is included in the GitHub repository
• LICENSE.md has been updated to reflect terms and conditions of source code usage

pb-StarPhase v0.14.2

Fixed

• Fixed an issue where a CYP2D6 graph with no edges would lead to a panic during SVG graph visualization

pb-StarPhase v0.14.1

Changes

• Added a new output folder through the --output-debug option: cyp2d6_igv_custom. This folder contains an XML file describing an IGV session as well as the supporting data files to visualize full length alignments through the two constructed CYP2D6 haplotypes. For details, see the updated user docs.
• Released an updated database file: data/v0.14.1/pbstarphase_20240826.json.gz

Fixed

• Fixed some off-by-one errors in the coordinates for miscellaneous extra regions in CYP2D6. These caused tiny overlaps in the output BAM file described above, but otherwise did not have an impact on diplotyping accuracy.

pb-StarPhase v0.14.0

Changes

• HLA allele labeling has been updated to improve 4th-field accuracy: When two potential definitions are compared, we now restrict the initial comparison to only the shared regions of the two haplotype sequence definitions (this is often different, especially for DNA sequences). In the event of a tie, we revert to the full-length allele definitions.
• The HLA database configuration has been updated to include strand information for HLA genes. Defaults for HLA-A and HLA-B are set, so no database update is required. This modification will show in the next database release.
• HLA debug consensus outputs will now be output on the strand the gene is located to improve matching to IMGT/HLA sequences. For example, HLA-A is already on the forward strand so no change will be made. In contrast, HLA-B is on the reverse strand so the consensus sequences will be reverse complemented in the output FASTA file.
• Breaking change: CYP2D6 and the HLA genes now share a single debug BAM file through the --output-debug option: debug_consensus.bam
  • The previous debug file for CYP2D6, cyp2d6_consensus.bam, has been removed from the outputs. The mappings from this file have been moved into the new debug_consensus.bam file.
  • For both HLA genes, the BAM file contains alignments of the HLA consensus sequences and corresponding read sequences used to generate the consensus. Additionally, if the assigned haplotypes have DNA sequences in the database, those sequences are also aligned for comparison purposes.
  • Previously deprecated option --debug-hla-target has been repurposed to allow for specification of additional HLA haplotypes to get mapped in this debug BAM. As with the assigned haplotypes, these must have a DNA sequence in the database to get mapped.

pb-StarPhase v0.13.3

Fixed

• Replaced a panic with an error message when low coverage datasets fail to identify any CYP2D6 haplotypes to chain together. These will have a "NO_MATCH" diplotype in the results.
• Fixed a bug where duplicate consensus sequences in CYP2D6 could create a panic, duplicates are now flagged as FalseAlleles and ignored.

pb-Starphase v0.13.2

Fixed

• Adjusted the alignment parameters for HLA mapping to reduce errors caused by soft-clipping of alignments near the end of a haplotype
• Replaced a panic with an error message when variants are found in unexpected states of zygosity and phase (e.g., phased homozygous)
• Debug messages for HLA calling have been adjusted to improve log reviewability

pb-StarPhase v0.13.1

Changes

• For HLA genes, StarPhase would previously ignore any HLA allele definitions that were missing a DNA sequence in the database. StarPhase now allows these partial HLA allele definitions by default.
• A new option was added to enable the previous behavior: --hla-require-dna. If this option is enabled, any HLA allele definition that is missing a DNA sequence will be ignored and never reported in StarPhase outputs.

Fixed

• Fixed an issue where a CYP2D6 deletion allele (*5) could be reported on the same haplotype as another allele. While this is biologically possible (e.g., deletion of one *10 in a "*10x2" haplotype), it is not considered a valid star-allele at this time. This combination will still show up in the debug log files, but it will get filtered in final reporting. For example: a "*10+*5" haplotype will now get reported as "*10".

pb-StarPhase v0.13.0

v0.13.0

Changes

• The algorithm for HLA-A and HLA-B has been modified to use a consensus-based approach to solve the alleles, a simpler version of the algorithm for CYP2D6.
  • CLI options related to consensus generation now control both HLA and CYP2D6 calling. These have been moved into a separate category on the CLI labeled "Consensus (HLA and CYP2D6)".
  • In internal tests, these changes slightly improved the accuracy of 4th-field entries in the HLA calls (2nd- and 3rd-field were unaffected). Additionally, the approach significantly reduced compute time requirements, averaging ~10% of CPU time required for v0.12.0.
  • With this change, the --threads option does not provide any benefit to the current algorithms. It has been deprecated, but may be added again if future optimizations allow it.
  • The --max-error-rate default has been adjusted for comparison to just the reference allele for each HLA gene, with a new default of 0.07 (previously 0.05).
  • Previous option --min-allele-fraction for HLA has been removed. The consensus option --min-consensus-fraction is used instead.
• Added a new option, --output-debug, that will create a debug folder with multiple additional files that are primarily for debugging the results from HLA and CYP2D6 calling, but may be useful for researchers. This folder is subject to change as the underlying methods develop. Some of the initial files included:
  • consensus_{GENE}.fa - Contains the full consensus sequences generated for a given {GENE}. Currently, this is only for HLA genes and CYP2D6.
  • cyp2d6_consensus.bam - Contains mapped substrings from the reads that were used to generate CYP2D6 consensus sequences. The phase set tag (PS) indicates which consensus the sequence was a part of. Useful for visualizing how the consensus ran and whether there are potential errors.
  • cyp2d6_link_graph.svg - A graphical representation of the connections present between CYP2D6 consensus segments.
  • hla_debug.json - Contains the summary mapping information of each database entry to the generated HLA consensus sequences.

Fixed

• Fixed an issue with build where CPIC genes with no known chromosome would cause an error and exit. These entries are now ignored with a warning.
• Fixed an off-by-one error in the HLA gene region start coordinates. This has been corrected in the latest database release: data/v0.13.0/pbstarphase_20240730.json.gz

pb-StarPhase v0.12.0

v0.12.0

Changes

• Hard-coded coordinates (GRCh38) for HLA-A, HLA-B, and CYP2D6 calling have been moved into the database file. Prior database versions without this new config information will automatically load the previously hard-coded values. This configuration is provided for transparency and experimentation in other reference coordinate systems, we do not recommend or support changing the provided default values.
• Released an updated database with the updated config format and name pattern change: data/v0.12.0/pbstarphase_20240716.json.gz
• Changed CYP2D6 consensus merging component to merge on sub-alleles instead of core alleles.
  • To support the this change, consensuses with ambiguous CYP2D6 assignments (e.g. equal matches to "*4.001" and "*4.015") are labeled as unknown for the purpose of merging with similar consensus alleles prior to generating a final consensus set.
  • Internal tests showed this combination of changes led to increased sensitivity for sub-allele identification without injecting errors at the core allele or diplotype level.

pb-StarPhase v0.11.3

Fixed

• Updated the build mode to account for IMGT-HLA's new database format that was released with v3.57.0-alpha