ET0H's analysis pipeline for the ubiquitin sequencing data.

Ipython notebook is recommended for interacting with the data. Assuming you are in the parent directory of seq-analysis, use as follows:

import pandas as pd
aa_weighted = pd.read_csv("seq-analysis/data/aa_weighted.csv", index_col=0)
# convert to Multiindex
aa_weighted.set_index(["days", "amino acids", "positions"], append=True, inplace=True)
# getting only the Et0H weighted mean slope data and have the days in 
# separate columns
aa_weighted.loc["Et0H", "weighted mean slope"].unstack("days")

We're using Anaconda to manage all packages. First install Miniconda from http://conda.pydata.org/miniconda.html. Either Python 2 or Python 3 will work, because we will be creating a virtual environment. It should add the conda executable to your path, if which conda give you nothing then make sure to add conda. Then run

conda create -n seqanalysis -y python=3 seaborn ipython-notebook biopython scipy scikit-learn
source activate seqanalysis
ipython notebook

The raw processed data is called lanes_new_combined.fastq.h5 and it's located in the et0h/data folder on the derisilab105 server.

There are two main files: seq_analysis.py and demultiplex_fastq.py. The latter takes a raw fastq file and demultiplexes it. The former actually goes through and processes the data. Most of the interesting stuff is there.