Hands-on workshop material for using Jupyter Notebooks on the CloudOS platform

After you have forked the repository, <USER_aka_you>/jax-jupyter, click on and click SSH to change the url.

conda config --add channels defaults
conda config --add channels bioconda
conda config --add channels conda-forge

WIP: Create markdown table for the most handy ones, here is a good place to start: http://maxmelnick.com/2016/04/19/python-beginner-tips-and-tricks.html

Pandas is an excellent python library that simplifies basic data handling tasks for dataframes. In this part, we will explore a well known dataset, iris, and we will learn how to perform basic data cleaning and handling operations using pandas utilities along with python build in functions. We will use Pandas to read tabular files in a dataframe, inspect the dimensions, variable values and number of observations, create subsets of our data based on column and row filtering criteria and retrieve summary statistics for our dataset. In the end of this part, we will have created a basic exporatory data analysis workflow for performing a minimal data quality control routine.

For the second part of the workshop, we will switch to R to explore the great functionalities of ggplot2 and related R packages for publication ready plots. We will first create a suitable input table with helper variable to facilitate the plotting. We will explore the ggplot layers and customise different aesthetics to create publication ready plots. Finally, we will add interactivity to our plots by using the plotly library. We will also explore two extremely handy R packages, ggpubr and

What does it take to submit a fully reproducible research piece? Notebooks, Jupyter Notebooks or Rmarkdown documents for example, have been proposed as a promising solution to accompany and bundle together in a reproducible report all the supplementary data of an scientific analysis that get published. Enabling other researchers reproduce our published work, offers the possibility of Postpublication Peer Review. During the Open Data Day 2017, hosted by SPARC and the NIH, within the two-day hackathon organised, an attempt to analyse all Jupyter notebooks mentioned in PubMed Central started, with many contributors from the open source community, including Daniel Mietchen. Thanks to this initiative, a collection of metadata for publications including Jupyter Notebooks was created. At the time, approximately 100 publications from EuropePMC, which included Jupyter Notebooks to assist with reproducibility of their analysis were documented. Our goal for this part of the workshop is to select a publication with a Jupyter Notebook, and try to reproduce the analysis. The objective of this part is to realise the minimal requirements of a reproducible workflow and hopefully explore the literate programming workflow of the publication authors. The repository where this effort was coordinated can be found here: https://github.com/sparcopen/open-research-doathon. A relevant JupyterCon talk by Daniel Mietchen can be also found here: https://youtu.be/Via7gBrjxHI

Moving on to our last part of our tutorial, we will learn how to use Integrative Genomics Viewer (IGV) on the CloudOS platform and easily connect data from your cloud provider. Contrary to the usual configuration required to use IGV, there are no prerequisites or installation steps to use IGV on CloudOS, nor the need to copy data. To access data we will simply link the data from our cloud storage provider and continue with the IGV analysis.