Using the Research Environment for clinical diagnostic discovery, January 2025¶
The rich phenotypic and medical history data, coupled with whole genome sequences available in the GEL Research Environment provides a unique opportunity for diagnostic discovery. In this training session we will take you through the data available in the GEL RE, including results of our own bioinformatic diagnostic analysis, and how you can access these, as well as tools available to filter the data, carry out your own analyses and validate your results. As part of the training, we will show you how to submit your genetic diagnoses to the clinical Genomic Medicine Service or contact clinicians, leading to direct clinical application of your work.
This training is aimed at clinical geneticists and does not require any coding skills. You will have chance to ask questions of our clinical team, who assess submitted diagnoses, and see what happens to your submissions.
Timetable¶
13.30 Welcome and introduction
13.35 GEL ingestion of rare disease participants
13.45 Identifying participants who need a diagnosis
13.55 Finding results of GEL analysis
14.05 Exploring variants in IVA
14.15 Validate your diagnosis
14.25 Find and compare other participants with the same variant
14.35 Submit your diagnosis and/or contact clinicians
14.45 Getting help and questions
Learning objectives¶
After this training you will know:
- How GEL analyses new rare disease genomes and where to find the results of these analyses
- How to filter, analyse and validate variants in a participant of interest
- How to submit diagnoses to the GMS
Target audience¶
This training is aimed at researchers:
- Are working in the Genomics England TRE
- Are clinical geneticists trying to do diagnostic discovery
- Do not necessarily have any coding skills
Date¶
14th January 2025
Materials¶
You can access the redacted slides and video below. All sensitive data has been censored.
Slides¶
Video¶
Give us feedback on this tutorial
Q&A¶
Q&A
I want to separate phenotypes by history of anorexia. I’ve used participant explorer to build my cohort of around 155 participants. I want to know 3 things.
- Can use IVA to only look at variants of my 155 participants. Not targeting a specific gene, or looking at them one by one. Can I group the 155 participants as one group and use IVA to analyse all of their variants?
- If I can’t do this on IVA, can I do this using aggv2?
- For aggv2, is there any way of using build 37 as well as 38? Like lifting over to build 38?
live answered
How do I find if a patient has any tier 1 or tier 2 variants, if the clinical report says unsolved?
live answered
Is exomiser restricted to exons of genes in a panel app?
Exomiser is independent from the Genomics England tiering algorithm and does not utilise PanelApp.
Is it correct that the submitted diagnostic discovery means that patient has been solved?
The submitted_diagnostic_discovery table includes candidate diagnoses that have been identified through research analysis. These variants have been reviewed at Genomics England and shared with the NHS laboratories. These candidate diagnoses are subsequently reviewed clinically by the NHS Genomic Laboratories according to clinical best practice guidelines.
Can you filter phenotype by ICD-10?
live answered
How reliable are the HPOs - were they submitted individually at time of recruitment?
live answered
Do you have to say Yes/No for the HPOs in participant explorer?
live answered
Can I identify a patient via a patient ID/ from their clinical report of the WGS findings?
The National Genomic Research Library is a de-identified database, and this is reflected in our participant information sheets and in turn their consent.
Is there an easy way to get a list of the genes that are different between different versions of each pannel?
live answered
Is there information about histological diagnosis (for example in high grade glioma patients) which can be used to filter participant ids, and in turn variants?
live answered
What’s the difference between population frequency and cohort alternate stats?
live answered
I understand it is de-identified, but as a clinician, if I want to look for variants in my patient who has a negative report, who do I contact/ how do I find out their patient ID in the NGRL?
live answered
Is there a prediction on when IVA will be made available for patients in the NHS releases?
live answered
Is there a way of identifying which patients were sequenced on the same run? In case I want to load individuals from the same run on IGV as a control to check for systematic errors
live answered
I would also like to know, seeing as participant explorer is not available for patients in the NHS release, how does one go about accessing medical history for those individuals?
live answered
Hi,
In the teiring table, you often find in the column for Penetrance that variants are marked as Complete or Incomplete. What does this mean?, and how different is the penetrance column from the Segregation column? Thanks
live answered
If I want to look through >100 participants for variants, will I need to do it individual by individual?
live answered