Skip to content

Running an Experiment

Victor Lin edited this page May 22, 2020 · 1 revision

It's important that every experimental result in serratus be 100% reproducible and all the data is freely available amongst contributors. Each notebook entry should be documented such that a reasonable bioinformatician can repeat you work exactly.

All results are documented in a shared electronic notebook: serratus/notebook

To assist in this, we suggest using Jupyter Notebook to document all the code in running an experiment. Alternatively you can use markdown or other ways where code and output can be distinguished.

Experiment organization

Notebook entries naming and file convention. (YYMMDD is start date).

# Notebook Entry
serratus/notebook/YYMMDD_experiment_name.ipyb

# Scripts, small data files (>250 kb) and plots from experiment
serratus/notebook/YYMMDD/<filename>.Rscript
serratus/notebook/YYMMDD/<filename>.png
serratus/notebook/YYMMDD/<filename>.csv
...

# Large data files are stored on S3
s3://serratus-public/notebook/YYMMDD/bam/aligned.bam
s3://serratus-public/notebook/YYMMDD/large_data.Rdata
...

Experiment Template

Copy the experiment template to start your experiment.

cp notebook/200401_template.ipynb notebook/200420_my_experiment_title.ipynb

Depositing Data

Please contact us to be granted permission to deposit data

Update the notebook Table of Contents

Upon completion, add your experiment and a one line description of it to the Notebook Table of Contents.

Clone this wiki locally