microarray

I was recently asked how to get up-to-date annotations of microarrays. One way is to use a current version of cDNAs and the probe sequences supplied by the manufacturer. The following examples were used in conjunction with Agilent 44K microarrays (Physcomitrella and rice to be specific). I may add an Affymetrix example as well.

some preparation steps (align the probes to cDNA sequences which you are interested in)

The pre-processor script relies on a working Bowtie installation. The binaries are assumed to be located in your PATH. To see all options of the python script type:

python prepareMicroarrayProbes.py

Download and unpack the pre-process test data. The file called Agilent-017743_GPL14653_spotSequences.txt corresponds to the seqTable.txt and PpatensV6_filtered_cosmoss_mRNA.fasta to cDNA.fasta.

build the bowtie index

To align the probes to the cDNAs of interest, one needs to build a bowtie index first. Note that the locus ID (so the ID that gets the expression value) corresponds to the first field after the arrow (>) in the fasta file (split using space character).

python prepareMicroarrayProbes.py BUILD cDNA.fasta cDNA_index

reformat the sequence table

Probe sequences are frequently stored in tabular form. This needs to be changed into a fasta file.

python prepareMicroarrayProbes.py TABTOFASTA seqTable.txt 1 1 2 probes.fasta

align sequences

To align the probes to the cDNAs of interest:

python prepareMicroarrayProbes.py ALIGN cDNA_index probes.fasta unaligned.txt aligned.txt

extract the mappings

Extract the probe name to locus ID mappings (and vice versa):

python prepareMicroarrayProbes.py EXTRACT aligned.txt probeNameToID.txt IDtoProbeName.txt

processing in R, 44K agilent single-channel arrays

download and unpack the rice test data and see agilentSingleChannelExample.R

processing in R, 44K agilent dual-channel arrays

download and unpack the Physcomitrella test data and see agilentDualChannelExample.R