TLDR - Check out this github repo for a (still really wordy) example: polyglot_jupyter_example

If you're anything like me, and there are probably tens of you out there, you enjoy working in multiple programming languages for your bioinformatics/data science work. Perhaps you love the tidyverse R ecosystem for data manipulation …

Continue reading

TLDR - Check out this github repo for the full example: https://github.com/groverj3/minimal_nextflow_samplesheet_example

I recently wrote an article regarding some of my opinions on bioinformatics workflow design. I've written workflows in several languages over the years, but at this point it seems that Nextflow has become something of …

Continue reading

How do you communicate results of an analysis? What tools do you use? Scientists that work in the wet lab are accustomed to firing up excel or some instrument-specific software and working with their own data. For genomics or other types of experiments in biology that result in large datasets …

Continue reading

Since I was in grad school I've been writing bioinformatics workflows. Usually to process NGS data. The concept of a workflow is simple, and not limited to the domain of bioinformatics. However, a workflow (aka "pipeline") used to analyze data from next generation sequencing (again, will it ever be "current …

Continue reading

One of the, if not the, most common downstream analysis task I'm asked to perform on RNAseq data is to generate the venerable "Volcano Plot." These are kind of the bioinformatics equivalent of saying "Hey! Look how much data I have!" Regardless, they are a pretty good way to quickly …

Continue reading

In contrast to the zen of python there are actually many ways to handle sequence data in Python. There are several packages on PyPI that provide parsers for sequence formats like .fastq and .fasta. I've never bothered with these, including the oft-used Biopython. I vaguely remembered Biopython being slower than …

Continue reading

Over the past few years the use of workflow managers in genomics and bioinformatics has grown greatly. This is a great thing for the field and adds to our ability to perform reproducible analyses, especially for pipelines with many steps. These are common in bioinformatics, but prior to the use …

Continue reading

Bioinformatic analyses often require lengthy workflows or pipleines, where the output of program A feeds into program B, and so on. These programs may also not output their results in a format which is convenient to use in the subsequent steps, requiring writing a conversion script, or piping its output …

Continue reading

Working with large amounts of tabular data is a daily occurance for both bioinformaticians and data scientists. There's a lot the two groups can learn from each other (great future post material). However, I recently ran into a situation that I was sure had to be relatively common. Apparently it …

Continue reading

When analyzing RNAseq you're faced with many possible analysis pipelines. The biggest decision you need to make is what the purpose of your experiment is. I will make the assumption that most of the time people want to determine which genes are differentially expressed between two samples, genotypes, conditions, etc …

Continue reading