standard analyses and code sharing project update
play

Standard Analyses and Code Sharing Project Update (PhUSE - PowerPoint PPT Presentation

Standard Analyses and Code Sharing Project Update (PhUSE Computational Science Working Group) PhUSE Webinar 16 November 2016 Working Group Vision/Goals Leverage crowd-sourcing to improve the content and implementa6on of analyses for medical


  1. Standard Analyses and Code Sharing – Project Update (PhUSE Computational Science Working Group) PhUSE Webinar 16 November 2016

  2. Working Group Vision/Goals • Leverage crowd-sourcing to improve the content and implementa6on of analyses for medical research, leading to be;er data interpreta6ons and increased efficiency in the clinical drug development and review processes. – Establish and maintain a publicly available repository for storing program code to be used as analy6cal tools for medical research. – Where gaps exist, develop recommenda6ons for analyses and displays in areas that could benefit from crowd-sourcing. – Where gaps exist, develop code for recommended analyses and displays that could benefit from crowd-sourcing (to reside in the repository).

  3. Vision: Fill the Gap on Analysis Standards Tables, Data Clinical Data Figures Trial Observed Analysis Collec6on Flow and Design Datasets Datasets Systems Lis6ngs Industry No TFL Stds PRM SDTM ADaM CDASH Standards Exist Alignment CFAST (Therapeu6c Areas)

  4. Vision: Shared Reusable Code Library (Script Repository) FDA Industry Academia Repository: Standard Qualified SAS Scripts Cloud R Studio Whitepaper Contributed Scripts Scripts

  5. Focus Areas • Three Focus Areas, 6 acBve projects Script Repository Script Discovery and Acquisi6on (SDA, (4 projects) Rebeka Revis, Michael Man) Repository Content and Delivery (RCD, Analyses and Display Gustav Bernard, Andrew Miskell) White Papers Repository Governance and (ADW, Mary Nilsson) Infrastructure (RGI, Mike Carniello, Hanming Tu) Communica6on, Test Data Factory (TDF, Peter Promo6on, Educa6on Schaefer; TBD) (CPE, Jared Slain and Wendy Dobson)

  6. Script Discovery and Acquisition (SDA) • JumpStart scripts updates – Received scripts that are part of FDA’s JumpStart service • A couple in 2015, addi6onal scripts early 2016 – First draa of scripts have been tested – Scripts are currently being updated to create a more user-friendly package • Should be done by the end of 2016 – The working group will need script reviewers/testers once the updates are completed – Outlining differences between JumpStart output and white papers is in progress – Instruc6onal presenta6on at 2016 PhUSE Annual Conference • Non-clinical scripts, scripts related to data handling have also been contributed

  7. Repository Content and Delivery (RCD) • Scripts exist in the repository – Several scripts at various stages (5 Script-athons, plus addi6onal work by the project team) • Focused on crea6ng scripts associated with the displays in the white papers, star6ng with ADaM data • The specifica6ons for the scripts include assump6ons and required ADaM variables • Star6ng with CDISC pilot data

  8. Repository Governance and Infrastructure (RGI) • Script Repository has been expanded – MIT license chosen: h;ps://github.com/phuse-org/phuse-scripts/blob/master/LICENSE.md – Qualifica6on guidelines developed: h;p://www.phusewiki.org/wiki/index.php?6tle=WG5_Project_02#Qualifica6on_Process – User-friendly front end developed: h;ps://github.com/phuse-org/phuse-scripts/wiki/Simple-Index – Repository has been updated to have a place to store scripts developed by other groups: • Non-clinical: h;ps://github.com/phuse-org/phuse-scripts/tree/master/contributed/Nonclinical • Data Handle: h;ps://github.com/phuse-org/phuse-scripts/tree/master/lang/SAS/datahandle – More people have joined the development team (28 now): h;ps://github.com/orgs/phuse-org/people

  9. Test Data Factory (TDF) Background: • – Several PhUSE projects develop scripts/concepts that need to be tested with trial datasets – Use of CDISC pilot datasets and/or datasets from team members for tes6ng isn’t sufficient in many cases Approach and Objec6ve: • – Determine test dataset requirements (like # of subjects, observa6on parameters, etc.) – Implement scripts that create ‘simulated’ SDTM and/or ADaM data sets – Publish datasets, and if possible, user-configurable scripts Team: • – Project co-lead: Peter Schaefer (pschaefer@vca-plus.com) Status: • – Project team mee6ngs will start – Working on scope specifica6on and priori6es – Need addi6onal members

  10. A Proposed New Project - Best Practices on Data Collection Instructions • Collec6on instruc6ons – Adverse events, concomitant medica6ons, study disposi6on, reasons for permanent discon6nua6on from study treatment – Will poten6ally lead to increased efficiency (site personnel won’t have to learn as many collec6on methods) and fewer queries (less confusion by site personnel), – Will enable more standardiza6on in data, analyses, displays – Project lead iden6fied (Todd Bazin) – If approved, will be managed by the Op6mizing the Use of Data Standards Working Group

  11. Analyses and Display White Papers (ADW) • 5 White Papers finalized How to find final white papers: Go to – Vital Signs, ECGs, Labs - Central Tendency www.phuse.eu, • Finalized in October 2013, Lead – Mary Nilsson Click on CS Working Groups, – Non-Compartmental Pharmacokine6cs Click on CS • Finalized in March 2014, Lead - Francois Vandenhende Deliverables – Demographics, Disposi6on, and Medica6ons Catalog • Finalized October 2014, Lead – Simin Baygani – Vital Signs, ECGs, Labs – Outliers and Shias • Finalized in September 2015, Lead – Wei Wang – QT Studies • Finalized in March 2016, Lead – Christos Stylianou

  12. ADW - Accomplishments (cont’d) • AE White Paper in progress • 2 nd draa review cycle completed • Expected to be final by end of 2016 • Lead – Sheryl Treichel, Nhi Beasley, Mary Nilsson • h;p://www.phuse.eu/cs-deliverables-under-review.aspx • Upcoming white papers for public review – Hepatotoxicity • Working on Draa 1 for public review • Should be ready before the March 2017 CSS • Lead – Terry Walsh

  13. Working Group Needs • More project team members • Increased par6cipa6on in white paper reviews – Help recruit! • Use/reference recommenda6ons in exis6ng final white papers! – Forward to any exis6ng standards groups – Link to white papers from Sta6s6cal Analysis Plans • Par6cipate in re-usable code development – Write, Test, Qualify, Review, Improve • Keep eyes open for exis6ng scripts that need a public home • Adver6se! Adver6se! Adver6se!

  14. How to Participate • Sign up for the PhUSE working group mailings – From phusewiki.org, click “Join a Working Group Now” – Standard Scripts Groups • CSS-WG-Standard-Scripts (En6re Working Group) • CSS-WG-Standard-Scripts-WhitePapers (White Paper Project Team) • CSS-WG-Standard-Scripts-Platorm (Script Repository) • CSS-WG-SS-WhitePaperReviewers (No6fied when a white paper is ready for review) • See wiki pages for each of the projects (www.phusewiki.org)

Recommend


More recommend