<h1>Pilot Project for Medical Librarians to Help Document and Preserve Scientific Code at their Institutions!</h1>
<h2>This pilot will run from February to April, 2018</h2>
<h3>About the National Network of Libraries of Medicine (NNLM) and NNCO</h3>
<p>NNCO is the National Network Coordinating Office which leads the NNLM.</p>
+ Scientific Reproducibility
+ People write new code themselves
+ 1-2 years - code becomes dead
Identification of projects
+ Personal contacts
+ ~4 projects/institution
+ 1 hour per week? (total of 8 hours)
++ Is that feasible?
Things to measure
+ Interaction between librarians and biomedical coders
+ Improvement in documentation
+ Increase in understanding of code by librarians
++ Code within environment
Levels of engagement
+ Naive librarians
++ No coding experience
++ Help with documentation
++ Help with advertising
++ Help with organizing manuscripts
++ Archiving and DOIs
+ Moderate librarians
++ Help with error handling
++ Getting people to work in more modular environments
+ Code-savvy librarians
++ Help with continuous integration
+++ e.g. <a href="https://github.com/travis-ci/travis-ci" rel="nofollow">https://github.com/travis-ci/travis-ci</a>
Integration into existing data catalog
+ Being set up at several network libraries
+ Metadata does not necessarily exist.
Making the library itself more external facing
NLM Data Science Capstone Project Participants
Are there things where "best practices" are higher throughput
What does interaction look like?
+ Hackathon-like meetings?
+ User groups?
+ Code review?
+ Creation of services?
<a href="https://datashare.ucsf.edu/stash" rel="nofollow">https://datashare.ucsf.edu/stash</a>
+ California Digital Library
<a href="https://omictools.com/" rel="nofollow">https://omictools.com/</a>
<a href="https://github.com/seandavi/awesome-single-cell" rel="nofollow">https://github.com/seandavi/awesome-single-cell</a>
<a href="http://search.hsls.pitt.edu" rel="nofollow">search.hsls.pitt.edu</a></p>