Assessment tools overview
Updated over a week ago

Summary

When searching a variant the Assessment tab display a Summary section on top.

The summary contains the highlights on Franklin's interpretation data

Region Viewer

Franklin Region Viewer shows the amino acid string and variant coding sequence, marked with the classified variants and the domains or Curated Submissions and Dosage Sensitivity (for SV) in the variant region.

Hovering on the submissions markers, which present the variant classification by colours and the clinical evidence count by size(for snp), reveal the variant details.

Region Viewer for SV

Clinical evidence

The clinical evidence section lists the evidence and rated submissions from Clinvar, UniProt and also Franklin Community classifications.

References

When searching for SV variants you can find the references section with Franklin Community and Clinvar evidences

Please go to the "Clinical evidence for SV" article for details regarding evidence supporting the clinical relevance of SV variants.

Population Frequencies

Franklin provides an aggregated frequency based upon multiple public genetic frequencies database - 1000 genomes, Exome Aggregation Consortium (ExAC), Exome Sequencing Project (ESP 6500), UK10K, gnomAD (Exome and Genome) and more.

When expanding to a more detailed view, you can find the Subpopulation, Allele Frequency, Allele Number and Homozygotes data of each database.

Occurrences

Franklin's SV occurrence aggregates information about the distribution of the SV across available population databases.

Variants are called with differing methodologies (see below). As such, aggregation across databases allows for a more meaningful comparison of results.

Variants are considered "identical" if they are of the same type and share at least 50% reciprocal overlap in genomic location.

The occurrence graph displays the variant gnomAD frequency and DGV count.

Samples appearing more than once in the databases are counted only once in the aggregated occurrence.

Predictions

For each variant multiple in-silico prediction tools scores are provided, such as

SIFT, FATHMM, DANN, MetaLR, REVEL, MutationAssessor, PolyPhen-2, MutationTaster, SpliceAI, dbscSNV, GERP, GenoCanyon and fitCons.

Tools selection is done according to the type of the variants e.g. missense, splice site etc.

In addition these scores are aggregated into a single score using unique machine learning algorithms, in order to provide an easy filtering of important variants.

Please read the “Franklin’s prediction tools and score range” article for a deeper understanding of this subject.

When expanding this section you can find more detailed data under these categories- Aggregated, Functional Coding, Splice Altering, Conservation and .Functional Whole Genome.

Transcripts

For each variant Franklin List all transcripts from both ensemble and refseq.

The transcripts of the genes are displayed with their in silico predictions under different sections-

Missense, Synonymous, Splice Region ,Upstream, Downstream and Intronic
The Canonical transcripts are tagged.

__________________________________________________________________

More Assessment tools available for case variants

Confidence

Each variant is given a confidence score which indicates the probability of the variant being a false variant. This score is based on multiple metrics such as read depth, alignment quality metrics, number of variant callers that detected this variant and more.

Please look for “Franklin’s confidence criteria” article for more information.

In the assessments tools tab you can find the summary of the snp confidence values

For CNVs called by Genoox CNV caller, additional confidence parameters are calculated and graphically presented, including control samples illustration.
Parameter values are calculated for each exon in the chosen range, and aggregated for the whole variant - in the widget tabs.

Parameters description:

  • Median Depth - median depth of each exon.

  • Prediction Score - measures the quality of the exon prediction model. It is based on the ratio of samples in model validation cohort whose exon copy number was properly predicted. Typically values are between 0.7 and 1.0 where scores closer to 1.0 represent exons for which calling quality is higher. Value of 0 means there is no valid prediction model for the exon and calling is not performed for this exon.

  • Predicted Copy Number - the exon-level copy number predicted by the model. e.g. for heterozygous deletion is non-sex chromosomes, the Predicted Copy Number should be close to 1 and duplication would have Predicted Copy Number around 3 (or above).

Internal Frequency /Occurrence

Franklin calculates and maintains a personalized frequency metric for each organization based on the samples cohort which are stored in the organization repository.

This allows comparison against organizational data, variant internal frequency and knowing in which

sample the variant was previously reported.

When reviewing SNP and Indel variants the internal frequency metric is displayed

When reviewing SV variants the internal Occurrence metric is displayed, variants are considered "identical" if they are of the same type and share at least 50% reciprocal overlap in genomic location.

By expanding the variant tile, a table with more detailed information regarding variant frequencies will appear - here you could also locate the number of times the variant was found in your sample repository

By clicking on the number of samples, the Variant Assesment page tab will open, and a full list of the samples in which the variant appeared, as-well as additional information such as zygosity will appear

Genoox Global Frequency

Franklin provides additional unique frequency data, this is the variant frequency in all the cases Franklin’s users have uploaded.

Franklin invites the user to contact other members, who have the same variant in their cases, so they could contribute in solving the case.

Suspected Compound Variants

Franklin lists all the VUS/ Pathogenic /Likely Pathogenic suspected compound heterozygous variants in the same gene

Gene Panels

Franklin public panels are pre defined gene sets, for example “ACMG Incidental Findings” panel.

Franklin+ users can define their own curated genes sets in order to filter the case variants to these specific genes.

Gene Coverage

Franklin displays Gene Coverage details for FASTQ files that are available for upload only for Franklin+ users. In addition to Franklin’s coverage page, of the full sample coverage data, Franklin lists the gene’s areas with coverage <20 and with known pathogenic variants in the variant assets tab.

Sequence Browser

Franklin enables full access to the raw sequenced data in a graphical

format. the IGV browser, which demonstrates the specific reads and how the variants are reflected allows users to validate the quality or the variant source, verify regions

with low quality, view alignment details, and identify specific variants in the coverage

area with a lot of “noise”.

Please note that the sequence browser will only display data for FASTQ files that are available for upload only for Franklin+ users.

The viewer is accessible also directly from the variant tile by clicking on its chromosome location.

To download a BAM file:Click on the ‘BAM URL’ tab on the bottom-left corner (this step copies the URL link)Paste URL link in a new browser tab.The download will begin automatically.

Custom Annotation (np)

Franklin+ users can add their own additional annotated data integrated with Franklin interpretation data.

Gene Summary

The last section in the assets tools tab is the gene summary informatio.

Still have questions? Reach out to our Support Team, they'll be happy to help!

Did this answer your question?