ngs-bits

cppREST is a C++ library designed for implementing HTTP API servers. It leverages the QT framework for socket management.

Key characteristics:

• Protocols: Supports both HTTP and HTTPS (note: HTTPS is recommended for production environments; HTTP should be avoided in production).
• Concurrency Model: The library follows a thread-per-request model, where a new thread is spawned for every incoming request.

The ngs-bits repository provides a comprehensive suite of short-read and long-read sequencing tools designed for NGS-based diagnostics. The tools are categorized by the type of genomic data they process (BAM, BED, FASTQ, VCF, etc.) and their specific functional purpose (QC, Gene handling, Phenotype handling, etc.).

Note on NGSD Dependency: Many advanced tools (e.g., BedAnnotateGenes, PhenotypesToGenes, VariantRanking) require the NGSD (Next Generation Sequencing Database), which contains gene, transcript, and exon data. Installation instructions for NGSD are located in doc/install_ngsd.md.

SampleIdentity is a tool used to identify if different sequencing datasets (WGS, WES, lrGS, or RNA) originate from the same patient. It achieves this by calculating the identity of 75 specific SNPs that are typically well-covered across these sequencing types.

Key Constraints & Usage Notes:

• Genome Support: This tool works for HG38 only.
• Scope: It is designed strictly for checking sample identity. It cannot determine the relatedness of samples (e.g., kinship) and does not work for sequencing panels.
• Efficiency: It is optimized to be faster and more memory-efficient than the SampleSimilarity tool.
• Input Requirements: If providing multiple BAM/CRAM files, you must provide a text file containing one path per line via the -bams parameter.

# Example usage with a file list
SampleIdentity -bams input_files.txt -out results.tsv -min_identity 98

Because the burden test evaluates variants across many samples for every gene, it is computationally intensive.

• Gene Limit: It is recommended to select no more than 200 genes to avoid extremely long processing times.
• Splice Regions: By default, the test uses variants in the exon region plus or minus 20 bases (the splice region). You can adjust this splice region size within the burden test dialog.

When performing a trio analysis, the system calls Uniparental Disomy (UPD) in addition to ROHs using the UpdHunter tool.

If UPDs are detected, the RE (Runs of Homozygosity/Exclusion) dialog will display a specific dialog containing the following information for each UPD:

• UPD coordinates: Genomic location.
• size_kb: UPD size in kilobases.
• size_markers: UPD size in SNVs.
• upd_markers: The number of markers indicating a UPD.
• source: The source of the UPD markers.
• het_percentage: The percentage of markers with a heterozygous genotype.
• q-score: The quality of the UPD call (Phred-scaled).

The SampleSimilarity tool calculates several metrics to measure the similarity between sequencing samples. The specific metrics available depend on the input mode (VCF, GSvar, or BAM):

• overlap_percent: The percentage of variants found in both samples. This ignores genotype and is only available in VCF and GSvar modes.
• correlation: The correlation of variant genotypes for variants found in both samples.
• ibs0: The percentage of variants with zero Identity By State (IBS), such as AA and CC. This is only available in BAM mode.
• ibs2: The percentage of variants with complete IBS, such as AA and AA. This is only available in BAM mode.

Note that there are no universal cutoff values for these metrics, as results are highly dependent on factors like ethnicity and enrichment kits.

RohHunter results can be used to identify potential consanguinity in patients. Large ROHs are strong indicators of parental consanguinity.

Key Indicator: Based on sequencing data distributions, a total ROH size sum larger than 75Mb is considered a strong indicator for parental consanguinity.

To prevent GSvar files from becoming excessively large (WGS VCFs can contain ~5 million variants), only variants meeting at least one of the following criteria are included in the GSvar file:

• VEP impact is HIGH, MODERATE, or LOW.
• Annotated as pathogenic or likely pathogenic in ClinVar or HGMD.
• Annotated as class 4, 5, or M in NGSD.
• gnomAD Allele Frequency (AF) $\le$ 2%.
• Located on chrMT (mitochondrial DNA).

Note: Common variants (AF > 2%) that are purely intronic or intergenic are excluded.

Splicing junctions detected during alignment are visualized as tracks in IGV. When hovering over or clicking a splicing event, the system displays the number of supporting reads and the splicing motif.

Splicing event classifications:

• known: Matches a known transcript.
• exon-truncation: The upstream or downstream exon is truncated (junction occurs inside an exon).
• exon-skip: At least one exon is skipped.
• exon-skip-truncation: A combination of exon truncation and skipping.
• intronic-or-overlap-no-tx: No supporting transcript found; junctions are inside an intronic region (at least one end), suggesting a novel exon.
• other: Unspecific events, such as those in intergenic regions.

The src/ directory contains several .pro files used to manage different components of the project, and .pri files used to share settings across these projects.

Qt Project Files (.pro)

• libs.pro: Manages code and tests for base libraries (e.g., cppCORE, cppNGD, cppNGSD).
• tools.pro: Manages code and tests for ngs-bits command-line tools.
• tools_gui.pro: Manages code for GUI tools, specifically GSvar.
• tools_server.pro: Manages code and tests for GSvarServer.

Project Include Files (.pri)

These files are included in .pro files to share configurations:

• base.pri: Contains base settings; included by all other .pri files.
• lib.pri: Settings for libraries.
• test.pri: Settings for tests.
• app_cli.pri: Settings for command-line tools.
• app_gui.pri: Settings for GUI tools.