Michael Grünstäudl (Gruenstaeudl), PhD

Researcher at the Freie Universität Berlin

Alignment Phy2Nex few-liner

Alignment file format conversion for the efficient – Part II

Today, I needed to convert a series of alignments, which were stored in the phylip format, into the common nexus format. The output DNA alignment hereby needed to be of sequential format (i.e., non-interleaved).

In February 2017, I had already written a few-liner to conduct the inverse conversion (nexus to phylip) and was, thus, surprised to find that the conversion from phylip to non-interleaved nexus did not work out of the box. Instead, a few more lines (and a little trick using StringIO()) were necessary to get this specific conversion to work.

#!/usr/bin/env python2.7

import os
import sys
from Bio import AlignIO
from Bio.Alphabet import IUPAC, Gapped
from Bio.Nexus import Nexus
from StringIO import StringIO

inFn = sys.argv[1]
outFn= os.path.splitext(inFn)[0]+".nex"

inp = open(inFn, 'rU')
outp = open(outFn, 'w')

alphabet = Gapped(IUPAC.ambiguous_dna)
aln = AlignIO.parse(inp, 'phylip-relaxed', alphabet=alphabet)

out_handle = StringIO()
AlignIO.write(aln, out_handle, 'nexus')

p = Nexus.Nexus()
p.write_nexus_data(outp, interleave=False)


And for those who wish to apply the above Python code (saved as „phy2nex.py„) to a collection of directories which contain a phylip-file each:

for dir in */; do python2 phy2nex.py $dir*.phy; done

Standardizing tRNA anticodon abbreviations in genome files

Better U than T (or the other way around)

Incongruence exists among the annotations of tRNAs in genomic files: Most authors use „U“ (for uracil) in the three-letter anticodon abbreviations of tRNA anntotations, but some use „T“ (for thymine). While I consider the usage of both letters as justified, I believe that this usage should be consistent within a single genome file.

To assist with such a standardization, I have written a small Python script (directly executable from the bash shell) that replaces all „U“s of an anticodon abbreviation with „T“s (or vice versa, if so desired). Simply change the file name from „tmp“ to the actual filename in the below script.

python - <<EOF
import re
with open("tmp", "r+") as f:

lines = f.read().splitlines()
outL = []
kw = "\ttrn" # Depending on file format, keyword can also be "=trn" (e.g., gff-files)
for l in lines:

if kw in l:

pos = [m.end()+2 for m in re.finditer(kw, l)]
for p in pos:

l = l[:p] + l[p:p+3].replace("U","T") + l[p+3:] # Replacement from U to T; switch letters if desired






Note: P.S. Is indenting in HTML supposed to be that clumsy, with using different < p > < /p > definitions for every indent level?

Teaching award 2017

Awarded annually to a lecturer in biology, chemistry and pharmacy

I recently received the teaching award for biology at the FU Berlin for 2017. Many thanks to those students who recommended me. I appreciate your praise!


New paper – Conserved plastid genome structure in early-diverging angiosperms

Not so different after all.

Together with two other scientists, I just published a new paper on the plastid genome structure of early-diverging angiosperms. We demonstrate that the plastid genomes of early-diverging angiosperms are much more conserved than previously considered.


Teaching in spring

In the lab with teachers-to-be and high-school students

One of the master-level courses I teach at the Freie Universität Berlin during spring semester, the NatLab Evolution, is geared towards a comprehensive teaching education for upcoming high-school teachers. During this course, we invite high-school students from across Berlin to spend a day in the lab and work with the Master’s students on biological experiments. Given the general responses, both sides seem to enjoy this arrangement a lot!

Michael Gruenstaeudl – Teaching – June 2017

Michael Gruenstaeudl – Teaching – June 2017









Ordering charsets within NEXUS files

Character sets for the orderly.

Defining character sets (charset) in NEXUS files can be an efficient way to annotated specific regions of a DNA or protein alignment. However, many software packages able to write NEXUS files (e.g., BioPython) do not save charsets in an ordered fashion, if multiple ones are present (i.e., that charset at alignment positions 100-250 were listed before charset at alignment positions 350-550). Instead, a typical NEXUS file written by Biopython looks something like this:

CharSet foo = 5-7;
CharSet bar = 3-8;
CharSet baz = 2-6;

The following lines of bash code order the charsets of the NEXUS by their start position.

sed '/BEGIN SETS;/q' infile > outfile
awk '/BEGIN SETS;/,/END;/' | tail -n +2 | head -n -1 | sort -t' ' -n -k4 >> outfile
sed -n -e '/BEGIN SETS;/,$p' infile | sed -n -e '/END;/,$p' >> outfile

Teaching the teachers II

Teaching the teachers II

Like last year, I held training labs for current high school teachers as part of the NatLab initiative. It was a wonderful experience and great to find out what today’s high school teachers wish to communicate to their students.

NatLab Lehrerfortbildung – Apr 2017

NatLab Lehrerfortbildung – Apr 2017

NatLab Lehrerfortbildung 2017

NatLab Lehrerfortbildung – Apr 2017











Alignment Nex2Phy few-liner

Alignment file format conversion for the efficient

Today, I needed to convert a series of alignments, which were stored in the common nexus format, into newick format. In order to do this efficiently, I wrote the following few-liner.

#!/usr/bin/env python2.7
import sys
from Bio import AlignIO
inFn = sys.argv[1]
inp = open(inFn, 'rU')
outp = open(inFn+'.phy', 'w')
aln = AlignIO.parse(inp, 'nexus')
AlignIO.write(aln, outp, 'phylip')

The above-code is very ordinary and great to have handy.

Teaching freshmen

Currently, I am teaching a class on botany and global biodiversity to freshman-level students. Although the total number of students signed up for my course is large, and thus I have to guide dozens of students through the course materials every day, I love academic teaching and I hope that the students enjoy the course as much as I do.

Michael Gruenstaeudl - Teaching - Jan 2016

Michael Gruenstaeudl – Teaching – Jan 2016


Michael Gruenstaeudl - Teaching - Jan 2016

Michael Gruenstaeudl – Teaching – Jan 2016























Using rPython to call Python in R

A simple example.

Calling Python from within R via the R-package rPython is fairly easy. However, very little documentation exists on this package, and some of the commands may appear quirky at first. Also, don’t confuse rPython with RPython (see capitals)! The paucity of written documentation on this package seems to scare away many biologists, who – if they are like me – prefer to work with well-documented code. But fear not!

Usage of rPython is straightforward with a bit of exercise. The following tutorial is intended to give a quick and simple example on how to work with rPython.

> library(rPython)
Loading required package: RJSONIO

I use the command python.exec() to execute basically any Python operation within R.

# Generating a dictionary in Python
> python.exec("capcities = {'sk':'Slovakia','de':'Germany',

I use the command python.get() to assign a new variable in R with a value from Python.

# Pulling the dictionary from Python into R
> abbrev_dict = python.get("capcities")
> abbrev_dict
sk         de         hu
"Slovakia"  "Germany"  "Hungary"

Note, however, that python.get() is very inflexible in what it can retrieve from Python.

> abbrevs = python.get("capcities.keys()")
Traceback (most recent call last):
TypeError: dict_keys(['de', 'sk', 'hu']) is not JSON serializable

Instead, I use:

> abbrevs = python.get("list(capcities.keys())")
> abbrevs
[1] "sk" "de" "hu"