Skip to content

The Report

UnCoVar automatically generates a HTML-based report for a detailed insight into the analysed patient or environmental samples.

Section 1: Overview

Report

An overview table of key information through several stages of the workflow. The table consists of the following columns:

  • Sample: The sample name as entered in the sample sheet (config/pep/samples.csv) before starting the workflow.
  • Eukaryota to Unclassified: Classification of the sample contents on the domain level. Especially interesting for samples created via shotgun sequencing. with amplicon tiled preparation)
  • Raw Reads: Number of reads in the raw input of the sample.
  • Trimmed Read: Number of reads after adapter removal.
  • Filtered Reads: Number of reads after removing sequences that could belong to the human host from which the sample was taken.
  • Largest Contig: The length of the largest sequences after de novo assembly
  • De Novo Sequence: The length of the de novo assembled sequence after reference guided scaffolding. This sequence has less bias towards the SARS-CoV-2 reference genome. RaGOO
  • Pseudo Sequence: The length of the pseudo assembled sequence. This sequence has a stronger bias towards the SARS-CoV-2 reference genome, as detected variants (on read level) are applied onto that reference. Only applicable for samples process on Illumina devices.
  • Consensus Sequence: The length of the consensus sequences. This sequence also has a bias towards the SARS-CoV-2 reference genomes. It is created by using neural networks applied on a pileup of individual sequencing reads against the reference genome Only applicable for samples process on Oxford Nanopore devices. called variants with Varlociraptor into the Wuhan-Hu-1 Reference NC_045512.2
  • Best Quality: Indicates which of the assemblies performed produces the highest quality sequence and also indicates failed QC.
  • Pango Lineage: Called lineage in Pango nomenclature.
  • WHO Label: Lineage name determined by the WHO.
  • VOC Mutations: Mutations that occur in Variants of Concern (VOC). These VOCs typically have: Increase in transmissibility or detrimental change in COVID-19 epidemiology; OR Increase in virulence or change in clinical disease presentation; OR Decrease in effectiveness of public health and social measures or available diagnostics, vaccines, therapeutics. Other Mutations: Other Mutations that occur in the sample. of specific concern (VOC) and other found variants

Section 2: Variant Call Details

VOC Similarity

Estimation of VOC similarity based on each sample's SNV profile, compared to all VOCs from covariants.org.

  • Mutations: All possible SNVs found in the catalog with all SNVs of the most similar VOC showing first.
  • Probability: A-posteriori probability that the observation of the mutation is true.
  • Frequency: Variant allele frequency (VAF). The percentage of sequencing reads matching a specific SNV divided by the overall coverage at that position.
  • The colmuns with the VOC names represent the Top 10 similar VOCs, based on the similarity of their SNV profile to the SNVs found in the sample. measure of similarity. To determine the degree of similarity between the sample and the VOCs, we performed the following scoring. Let \(n\) be the total number of variants and \(m\) be the number of VOCs. Let \(X\) be a binary matrix that relates variants with VOCs, namely, \(X_{i,j} = 1\) if and only if variant \(i\) is in \(VOC_j\), with \(i = 1,\ldots,n\) and \(j = 1,\ldots,m\). Let \(\theta_i\) be the latent allele frequency of variant \(i\) in the given sample and \(\hat\theta_i\) be the maximum a posteriori estimate of \(\theta_i\) as provided by Varlociraptor. Let \(p_i = Pr(\theta_i > 0 \mid D)\) be the posterior probability that the variant \(i\) is present in the sample (i.e., the probability that its latent allele frequency is greater than zero, given the data \(D\)). Then, the similarity of a given sample to \(VOC_j\) can be calculated as the Jaccard-like similarity score:
\[\frac{\sum_{i=1}^n p_i \cdot \hat\theta_i \cdot X_{i,j} + (1-p_i) \cdot (1-X_{i,j})}{n}\]

Rendered variant callings

Variant callings rendered with Oncoprint:

  • with high and moderate/low impact
  • on ORF/Protein level

Variant candidates are identified and called with Varlociraptor

Section 3: Sequencing Details

MultiQC report for overall quality control

  • General Stats
  • Sequence Counts
  • Sequence Quality Histograms
  • Per Sequence Quality Scores
  • Per Base Sequence Content
  • Per Sequence GC Content
  • Per Base N Content
  • Sequence Length Distribution
  • Sequence Duplication Levels
  • Overrepresented sequences by samples
  • Top overrepresented sequences
  • Adapter Content
  • Status Checks
  • Software Versions

Coverage of Reference Genome

Plot that is showing how well the reference genome is covered by the reads of each analyzed sample visualizing the depth (DP) values from SAMtools depth.

Section 4: Sequences

Quality Overview

Filter Overview

A table comparing identity and number of N bases for all samples for the reconstructed genomes UnCoVar generates with its two main assembly methods (de novo assembly + scaffolding and consensus sequence based on called variants).

Pangolin Call Overview

A table of lineage calls for all samples throughout several stages of the workflow, as determined by Pangolin.

These stages are:

  • Scaffolded Sequences: Contigs from the de novo assembly, ordered against the virus reference genome (default: Wuhan-Hu-1 Reference NC_045512.2) with RaGOO.
  • Polished Sequences: Scaffolded sequence polished by applying variants with an allele frequency of 100% (called by Varlociraptor at FDR 5% by default).
  • Masked Polished Sequences: Polished Sequences masked due to two criteria. See below for a more in-depth explanation of these two criteria.
  • Consensus Sequences: Using corrected reads and the virus reference genome (default: Wuhan-Hu-1 Reference NC_045512.2) a consensus sequence is generated using medaka. Consensus Sequences are only generated for Oxford Nanopore data.
  • Masked Consensus Sequences: Consensus Sequence masked due to low position coverage. See below for a more in-depth explanation of the criteria. Masked consensus Sequences are only generated for Oxford Nanopore data.
  • Pseudo Sequences: Sequences that are1 generated based on the virus reference genome (default: Wuhan-Hu-1 Reference NC_045512.2) and by applying variants with an allele frequency of 100% (called by Varlociraptor with a default FDR of 5%). Masked Pseudo Sequences are not generated, as the quality criteria below are implicit considered when creating a "normal" Pseudo Sequence. Pseudo sequences are only generated for Illumina and Ion Torrent data.
Quality Criteria
  1. Positions in the reconstructed genome that are covered by less than a certain amount (default: 20) reads are be masked with N.
  2. Informative positions (A, T, G, C) which are not supported by a certain percentage (default: 90%) of the aligned reads are masked by non-informative placeholders of the IUPAC nucleotide code. Not applicable for model-based basecall procedures (e.g. Oxford Nanopore).

Section 5: Variant Call Files

Downloadable .vcf files of the different stages for variant calling:

  • with high and moderate/low impact
  • on ORF/Protein level

Section 6: High-Quality Genomes

  • Multi-FASTA file, including the reconstructed genomes from samples, that passed the quality control
  • .csv file, including additional submission data for each sample that passed the quality control