CONFIDENTIAL 1 February 13, 2019 Eric Morrie, Associate Director of Data Management February 13, 2019 R Programming for Data Management
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CONFIDENTIAL 1February 13, 2019
Eric Morrie, Associate Director of Data Management
February 13, 2019
R Programming for Data Management
CONFIDENTIAL 3February 13, 2019
Emanate Life Sciences
We’re a data management consulting firm with extensive knowledge in the Life Sciences industry. Our mission is to help our clients bring products to market that improve people’s lives.
• We specializes in:
• Data Management Advisory
• Data Management Service Provider
• Contracting / Permanent Placement
Introduction
Eric Morrie
Associate Director of Data Management⎼ Oversee all clinical trial and genetic specimen
data used in diagnostic development.
Veracyte, Inc.⎼ Provide genomic diagnostics to minimize
procedure complexity and reduce disease uncertainty for patients with cancer.
The views expressed in this presentation are those of the author/presenter and should not be construed to represent the
views of Veracyte, Inc.
CONFIDENTIAL 6February 13, 2019
R Programming for Data Management
CONFIDENTIAL 7February 13, 2019
Why is R attractive?
• Scientific and Statistical Language
• Powerful Regression Capabilities
• Designed for Reproduceable Research
• Focused on Ad-hoc Analysis and Data Exploration
• Earlier Adoption of Latest Techniques
• Over 13,000 Extension Packages
• Free Open Source and Paid Support Available
CONFIDENTIAL 8February 13, 2019
R Packages
Input/Output Utilities
OpenXLSX – Read and write ExcelHaven – Read and write BDATSASxport – Read and write XPTODBC – access database sources
Hmisc – SAS utilities and statisticsKeyRing – Password excryptionMailR – Send emailHTMLTools – HTML generation
Data Transformation Graphing/Layout
Data.Table – Filter, Sort, Group, and MergeDplyr – Summarise and convertStringR – Text parsing and conversionLubridate – Date parsing and conversionReshape2 – Aggregation, split, pivot
Rmarkdown – Page LayoutsGGPlot2 – Graphs and figuresScales – Axis customizationShiny – Interactive ExplorationShinyDashboard – Metrics DashboardsHighCharter – Interactive Graphs
CONFIDENTIAL 9February 13, 2019
Can I use R in clinical trials?
• Second Most Preferred Language for Clinical Research
• FDA does not endorse or require any specific software1
• Used by FDA on a daily basis
• FDA contributes R packages
• R is validated for Clinical Trials3
CONFIDENTIAL 10February 13, 2019
Comparison To SAS
SAS R
Language High-level data manipulation language Lower-level object-oriented language
Data Handling File, SQL, or Web Services
Data Processing File-based processing In-memory processing
Statistical Capabilities Only vetted methods Leading edge techniques
Interface Command-line, local GUI, server
Outputs Formatted Reports Manuscript Ready Layouts
CONFIDENTIAL 11February 13, 2019
Programming Interface
• Jupyter Notebook⎼ Supports R, Python, and Scala
• R Tools for Microsoft Visual Studio⎼ Use R Alongside Microsoft Languages
• R Studio⎼ Dedicated to R⎼ Package management is built in⎼ Supports Markdown and Shiny
CONFIDENTIAL 12February 13, 2019
Data Transformations
• Read data from any source⎼ Including XPT and BDAT
• Derive and Impute new variables
• Assign categories to continuous data
• Pivot data, including text
• Apply conditionals (If-Then-Else)
• Summary Statistics
• Utilize custom functions
CONFIDENTIAL 13February 13, 2019
Graphs
• Graphing in R ensures reproducibility
• Output format must match rendering⎼ Use PDF for print outputs⎼ Use PNG for web outputs
• Extensions for any visualization⎼ GGPlot2⎼ Plotly⎼ SunburstR⎼ HighCharter (not for printed graphs)
CONFIDENTIAL 14February 13, 2019
Formatted Layouts
• R Markdown for Layouts
• Includes Cover Page and Table of Contents
• Change Page Layout Mid-document
• Custom headers and footers
• Include text, tables, and charts
• Output to Word, PDF, RTF, and more
• Perform Burst Distribution COHORT 1
COHORT 2
COHORT 3
TRANSCEND Study Metrics Page 3 of 7
CONFIDENTIAL 15February 13, 2019
Analytics
• Shiny provides ⎼ Dashboards⎼ Data exploration⎼ Data file upload⎼ Report download
• Web browser access
• Graphs refresh automatically
• Supports data selection and filtering
CONFIDENTIAL 16February 13, 2019
References
1.https://blog.revolutionanalytics.com/downloads/FDA-Janice-Brodsky-UseR-2012.pdf
2.https://blog.revolutionanalytics.com/2012/06/fda-r-ok.html
3.https://www.r-project.org/doc/R-FDA.pdf
4.https://channel9.msdn.com/Events/useR-international-R-User-conference/useR2016/Using-R-in-a-regulatory-environment-FDA-experiences
5.https://www.burtchworks.com/2018/07/16/2018-sas-r-or-python-survey-results-which-do-data-scientists-analytics-pros-prefer/
6.https://www.aridhia.com/blog/r-for-researchers-8-essential-cheatsheets-for-research-data-analysis/
7.https://www.rstudio.com/resources/cheatsheets/
CONFIDENTIAL 18February 13, 2019
R Demo
CONFIDENTIAL 21February 13, 2019
Questions & Answers
Eric Morrie
Associate Director of Data Management
CONFIDENTIAL 22February 13, 2019
Upcoming Event
Centralized Monitoring: The Role of Process and Technology!
Wednesday, March 20, 2019
Veeva Systems | Pleasanton CA
5:30-7:30 PM Event Time
www.emanatelifesciences.com
CONFIDENTIAL 23February 13, 2019
Speakers and Volunteers Needed
Speaking Opportunities
• Sensors, Wearable & Biomarkers
• Artificial Intelligence in Clinical Trials
• Improving Site-Study Activation & Performance
• Managing Outsourced Clinical Trials
• Implementing CDISC: CDASH, STDM, ADAM
• Endpoint Adjudication
• Data Management
Contact Information
Hajime Arnold
Terri Buller
Related Documents
