Reproducible research 2 What is reproducible research? • Have you heard about it? • How would you definite it? • Is it the same as replicability? • Is it important?
Reproducible research 3 Research • Science moves forward then discoveries are replicated and reproduced Implementing Reproducible Research by Stodden, Leish, Peng
Researchers need a new definition for replication that acknowledges variation in both the original study and in the replication study. Specifically, a study replicates if the data collected from the replication are drawn from the same distribution as the data from the original experiment. Multiple independent replications of the same study will be needed to definitively evaluate replication. Patil, Peng & Leek, Perspectives on Psychological Science, 2016 Reproducible research Key conclusions:
Reproducible research 11 Replication • Replication, the practice of independently implementing scientific experiments to validate specific findings, is the cornerstone of discovering scientific truth. Implementing Reproducible Research by Stodden, Leish, Peng
Reproducible research 14 Reproducibility: from back in 2006 • However, because of the time, expense, and opportunism of many current epidemiologic studies, it is often impossible to fully replicate their findings. An attainable minimum standard is "reproducibility," which calls for data sets and software to be made available for verifying published findings and conducting alternative analyses. Peng et al, Reproducible epidemiologic research, Am J Epidemiol., 2006
Reproducible research 15 Reproducibility • Reproducibility can be thought of as a different standard of validity from replication because it forgoes independent data collection and uses the methods and data collected by the original investigator. Implementing Reproducible Research by Stodden, Leish, Peng
Reproducible research 16 A bit more practical • The sharing of analytic data and computer codes uses to map those data into computational results is central to any comprehensive definition of reproducibility. Implementing Reproducible Research by Stodden, Leish, Peng
Reproducible research 17 Why is it important? • Except for the simplest of analyses, the computer code used to analyze a dataset is the only record that permits others to fully understand what a researcher has done. Implementing Reproducible Research by Stodden, Leish, Peng
Reproducible research 19 Together • Reproducibility is the ability to take the code and data from a previous publication, rerun the code and get the same results. Replicability is the ability to rerun an experiment and get “consistent” results with the original study using new data. Results that are not reproducible are hard to verify and results that do not replicate in new studies are harder to trust. https://simplystatistics.org/2017/03/02/rr-glossy/
Reproducible research 23 Reproducible documents • Have you ever had your code in one file, your description of the results in another file? • Ever made copy-paste mistakes? • What if you were asked to change some models or revise the document? • Was it easy to maintain?
Reproducible research 25 Reproducible documents in R • R Markdown is the easiest • It's based on Markdown: simple human readable syntax • You maintain a single file! It has the • code, • figures, • description of results. • It then creates a file in the format you want to share with others.
Reproducible research 30 Reproducible research can still be wrong! • Unfortunately, the mere reproducibility of computational results is insufficient to address the replication crisis because even a reproducible analysis can suffer from many problems—confounding from omitted variables, poor study design, missing data—that threaten the validity and useful interpretation of the results. Although improving the reproducibility of research may increase the rate at which flawed analyses are uncovered […] it does not change the fact that problematic research is conducted in the first place. Leek and Peng, PNAS, 2015
Bioinformatics 49 Aims • Organize data, allow new entries, data curation • Develop tools and resources that aid in the analysis of data • Use these tools to analyze the data and interpret the results in a biologically meaningful manner Luscombe et al., Methods Inf Med., 2011
AUCAGUCGAUCACCGAU transcription RNA translation protein ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT DNA M M M Bisulfite RNA ChIP Genome slide adapted from Alyssa Frazee
AUCAGUCGAUCACCGAU transcription RNA translation protein ACTGACCTAGATCAGTCGATCGATCGTATACGATTACAAAATCATCGGCAT DNA M M M RNA slide adapted from Alyssa Frazee
expression data for ~70,000 human samples GTEx N=9,962 TCGA N=11,284 SRA N=49,848 samples expression estimates gene exon junctions ERs slide adapted from Shannon Ellis
expression data for ~70,000 human samples Answer meaningful questions about human biology and expression GTEx N=9,962 TCGA N=11,284 SRA N=49,848 samples expression estimates gene exon junctions ERs slide adapted from Shannon Ellis
expression data for ~70,000 human samples samples phenotypes ? GTEx N=9,962 TCGA N=11,284 SRA N=49,848 samples expression estimates gene exon junctions ERs Answer meaningful questions about human biology and expression slide adapted from Shannon Ellis
Category Frequency F 95 female 2036 Female 51 M 77 male 1240 Male 141 Total 3640 Even when information is provided, it’s not always clear… sra_meta$Se x “1 Male, 2 Female”, “2 Male, 1 Female”, “3 Female”, “DK”, “male and female” “Male (note: ….)”, “missing”, “mixed”, “mixture”, “N/A”, “Not available”, “not applicable”, “not collected”, “not determined”, “pooled male and female”, “U”, “unknown”, “Unknown” slide adapted from Shannon Ellis
Goal : to accurately predict critical phenotype information for all samples in recount gene, exon, exon-exon junction and expressed region RNA-Seq data SRA Sequence Read Archive N=49,848 GTEx Genotype Tissue Expression Project N=9,662 divide samples build and optimize phenotype predictor training set predict phenotypes across SRA samples test accuracy of predictor predict phenotypes across samples in TCGA test set TCGA The Cancer Genome Atlas N=11,284 slide adapted from Shannon Ellis
Sex prediction is accurate across data sets Number of Regions 20 20 20 20 Number of Samples (N) 4,769 4,769 11,245 3,640 99.8% 99.6% 99.4% 88.5% slide adapted from Shannon Ellis
Number of Regions 589 589 589 589 589 Number of Samples (N) 4,769 4,769 613 6,579 8,951 97.3% 96.5% 91.0% 70.2% Prediction is more accurate in healthy tissue 50.6% slide adapted from Shannon Ellis
expression data for ~70,000 human samples samples phenotypes ? GTEx N=9,962 TCGA N=11,284 SRA N=49,848 samples expression estimates gene exon junctions ERs Answer meaningful questions about human biology and expression sex tissue M Blood F Heart F Liver slide adapted from Shannon Ellis
Collaborators The Leek Group Jeff Leek Shannon Ellis Hopkins Ben Langmead Chris Wilks Kai Kammers Kasper Hansen Margaret Taub OHSU Abhinav Nellore LIBD Andrew Jaffe Emily Burke Stephen Semick Carrie Wright Amanda Price Nina Rajpurohit Funding NIH R01 GM105705 NIH 1R21MH109956 CONACyT 351535 AWS in Education Seven Bridges IDIES SciServer
Bioinformatics 78 Staying Current in Bioinformatics & Genomics • “focused on applied methodology and study design rather than any particular phenotype, model system, disease, or specific method” • “a software implementation that’s well documented, actively supported, and performs well in fair benchmarks” http://www.gettinggeneticsdone.com/
lcolladotor
sshimizu2006
genomethica
fr0gger
doradora09
ngehlenborg
mansuy
wd15
yasslab
keyakkie
koro485
naokiuc
ykamit
notwaldorf
chriscoyier
jennybc
destraynor
marktimemedia
danielanewman
rocio
jonyablonski
robhawkes
productmarketing
erikaheidi
brettharned