Advancing Deep Phenotyping for Pediatric Cancer and Structural Birth Defects

By: Nicole Vasilevsky, Adam Resnick, Deanne Taylor, Allison Heath, Nomi Harris and Melissa Haendel

Following up on the successful Innovation across the Phenotypic Translational Divide Webinar, the Monarch Initiative and Kids First Data Resource Center members co-hosted a second webinar on September 02, 2020. The Part 2 webinar further highlighted work from six additional Kids First X01 investigators and included talks from Kids First DRC researchers Deanne Taylor, Adam Resnick, and Allison Heath (all from Children’s Hospital of Philadelphia (CHOP)). The talks were followed by brainstorming about phenotype data collection and analysis methods and challenges faced in implementing these workflows.

Topics from Kids First X01 researchers presented at “Innovation across the Phenotypic Translational Divide Part 2” webinar on September 02, 2020.

The structured discussion section focused on some key questions, to gain insight into the types of datasets the investigators have, some of the challenges they are facing when accessing or analyzing data in other studies, and tools that are currently being used, or tools that would help support their research. Highlights from the discussion session are below:

• Our researchers are studying a myriad of childhood conditions. These include malformations (such as brain malformations, craniofacial defects, including clefting), cancers (intracranial germ cell tumors, osteosarcomas), and syndromic diseases like CHARGE syndrome.
• Challenges in accessing and analyzing phenotypes in other studies include dependencies on data in a wide variety of formats (PDFs, tex documents, and other medical record formats), and the lack of standardized classification systems, as well as issues with harmonizing diagnoses.
• The preferred format for receiving clinical data is spreadsheets. Structured formats using FHIR standards and tools may present new opportunities.
• REDCap is the most commonly used tool to support phenotypic data efforts (e.g. for collection and annotation).
• The most commonly used Common Data Elements are gender and ethnicity for NIH reporting; however, Kids First investigators often do not use CDEs in their research.
• Researchers noted a need for the KFDRC to provide more information about standards that can be used to structure their data.
• Further collaboration is desired, where subgroups could meet to talk about specifics for common phenotypes between X01 teams.

Based on prior interest and synergies, we are planning another webinar in this series focusing on cross-species phenotyping, which will highlight work by X01 researchers on variant modeling efforts. This will inform how the Data Resource Center can support cross-species research questions and ultimately improve the clinical outcomes for affected children. Further information and the agenda will be available here soon.

To learn more about this initiative and find out about future meetings, please contact Nicole Vasilevsky (vasilevs@ohsu.edu) or Valerie Cotton (valerie.cotton@nih.gov).

More information:

Innovation across the Phenotypic Translational Divide Webinar Part 2:
When: September 02, 2020
Organizers: Nicole Vasilevsky, Valerie Cotton, Adam Resnick, Deanne Taylor, Allison Heath, and Melissa Haendel
Agenda: Available here
Slides: Available here

Funding:
This webinar was supported by an NIH BD2K conference grant “Forums for Integrative Phenomics” to Melissa Haendel and Peter Robinson: U13CA221044.