Tutorial
This tutorial shows how to use ClintLR to perform LIRICAL analysis of phenopackets using adjusted pretest probabilities.
Exomiser data
ClintLR uses the Exomiser variant database that is available
from the Exomiser FTP site. The versions current as of
this writing are 2302_hg19.zip (for the HG19/GRCh37 Genome Assembly) and 2302_hg38.zip (for the HG38/GRCh38 Genome Assembly).
Choose the appropriate genome build, and download and extract the corresponding file.
Test data
We will use a VCF file that is available in the Exomiser distribution file (e.g., exomiser-cli-13.2.0-distribution.zip),
which is available from the same FTP site. The file is called Pfeiffer.vcf
and can be found in the examples folder of the unzipped Exomiser distribution file. This file spikes a pathogenic FGFR3
variant into a public VCF file and can be used to test ClintLR.
ClintLR Analysis Workflow
1. Select diseases
The tree view shows the loaded Mondo ontology. Tree item icons indicate which terms are OMIM (black circle) or OMIMPS (red circle) terms. You can jump to specific terms in the ontology by using the Browse Mondo Terms or Browse OMIM Terms search fields. The Mondo term search field autocompletes the Mondo ontology disease name/label, and the OMIM term search field autocompletes the OMIM term ID. Information for the selected term is displayed in an HTML window to the right of the tree view (Figure 1B).
2. Apply pretest adjustments to selected diseases
Clinical intuition is incorporated into the LIRICAL algorithm by way of a slider in the ClintLR interface. The slider increases the pretest adjustment, hereafter referred to as a, values that get passed into LIRICAL for the selected term and its descendants. In the tree, an up arrow in the selected term’s icon indicates that a has been increased for that term. The a values can be reset to defaults for all terms by clicking the Reset pretest probability adjustment values button below the tree view.
Note: The tree icons do not update automatically. To see icon changes, the tree item must be expanded/collapsed/selected again.
The a values can be viewed in a separate interface (File -> Pretest Adjustment Map -> Show Map Interface). If desired, the a values can be saved as a separate text file (File -> Pretest Adjustment Map -> Save Map to File), which can then be imported later (File -> Pretest Adjustment Map -> Load Map from File).
3. Adjust LIRICAL parameters
ClintLR includes fields for various LIRICAL parameters (Figure 1A and 1B), including the likelihood ratio (LR) threshold, the pathogenicity threshold, minimum diagnosis count, whether to display all variants, and the name of the results output file.
Additional LIRICAL parameters are located in the LIRICAL tab of the Resources interface. The values for these parameters are preserved between sessions.
4. Load a phenopacket
A phenopacket needs to be loaded (Load Phenopacket button) for the LIRICAL analysis to run correctly. ClintLR supports both v1 and v2 Phenopacket Schema versions.
5. Optional: Load a VCF file
The LIRICAL analysis can be performed using only phenotypic information, or using both phenotypic + genotypic information by loading a VCF file into ClintLR (Load VCF button).
6. Run LIRICAL
The LIRICAL algorithm can be run directly from ClintLR using the Run LIRICAL button.
Note: There is currently no progress bar for the Run LIRICAL button, however the analysis is being run in the background.
Analysis results
Once the analysis is complete, the results will be written to a new HTML file (Figure 1C) and opened in a web browser. This HTML file is also saved locally in the output directory specified in the Resources interface. The top of the page shows some information about the input files and a list of observed and excluded HPO terms. The next section shows summarized representations of the top candidates.
Each row in the summary shows the rank, post-test probability, and name/ID of the disease. The row includes a sparkline representation of the phenotypic profiles of each candidate, with green bars indicating positive contributions and red bars indicating negative contributions to the diagnosis. The last bar represents the genotype likelihood ratio if LIRICAL was run with a VCF file. Hovering over the individual bars will show the name of the HPO term or gene, and all sparklines show the terms in the same order.
LIRICAL then presents a detailed analysis of each of the top candidates. The summary shows information about identified variants and the phenotypic profile. Hovering over the graphic shows information about the likelihood ratio and the type of the match.
The remaining part of the HTML output page contains information about the other top candidates and a list of all diseases analyzed. The bottom of the page includes explanations and documents the settings used for the analysis.