Release notes - July 17, 2023
over 1 year ago by Marko Marinkovic
Recently published apps
We published Giraffe-DeepVariant workflow 1.0, Cramino 0.9.7 and kyber 0.4.0 tools from the NanoPack2 toolkit, as well as Pisces 5.3.0.0 tool, PureCN NormalDB workflow 2.6.4, PureCN workflow 2.6.4, zUMIs 2.9.7 tool, and AlphaFold 2.3.2 tool. Here are the details:
- Giraffe-DeepVariant workflow 1.0 is a pipeline for calling small variants using the pangenome reference. The workflow starts with sequenced reads (FASTQs, CRAM). Reads are mapped to a pangenome with vg giraffe and pre-processed (e.g. indel realignment) before performing the variant calling step using DeepVariant.
- Cramino 0.9.7 is a quick QC tool intended for long-read sequencing. It takes a BAM/CRAM format alignment file and creates a QC report with mean coverage, number of reads, their mean and median length and sequence identity relative to the reference genome.
- kyber 0.4.0 creates a 600x600 pixel heatmap image of read length and read accuracy from input alignment file (BAM/CRAM format).
- Pisces 5.3.0.0 does variant calling from aligned amplicon sequencing data.
- PureCN NormalDB workflow 2.6.4 builds a normal database which is used for coverage normalization in PureCN workflow.
- PureCN workflow 2.6.4 estimates tumor purity and ploidy, copy number and loss of heterozygosity (LOH). Calculated purity and ploidy combinations are sorted by likelihood score. Copy number and LOH data are provided by both gene and genomic region. The steps in the workflow include: preparation of an interval file for further analysis, calculation of coverage for tumor and normal samples (optionally for additional tumors) and final calculation of purity, ploidy, copy number and LOH results.
- zUMIs 2.9.7 takes RNA-seq data with or without UMIs, STAR index files archive and annotation GTF file and analyzes the data as specified by the other input parameters.
- AlphaFold 2.3.2 is a machine-learning application which incorporates knowledge about physical and biological protein structure properties into a deep learning algorithm, and predicts protein structures with high accuracy.