We evaluated the expression of 14 222 genes among the studied tissues and for 12 331 genes we were able to assign a genomic position according to the NCBI mouse genome assembly (build 32, mapping 19 684 known LocusLink genes).
This allowed us to avoid any potential discrepancy caused by varying genomic distributions in earlier versions of the mouse genome assembly.
On the basis of a genomic contig (NT_026437) of a more recent mouse genome assembly (build 38), Wheeler et al. (2012) also published the chromosomal coordinates of the mouse RNases, which is consistent with our results.
The hidden Markov models were constructed using HMMer [ 52] from the coding sequences of the manually curated multiple sequence alignment of the mouse and rat Abp genomic DNA alignments (including novel predictions from the mm8 mouse genome assembly).
We mapped the genomic locations of all the L1 elements with the UCSC mouse genome assembly (March 2006, NCBI Build 35, version mm8) using BLAT analysis, with data accuracy greater than 0.98.
All reads were realigned to the NCBI m37 version of the mouse genome assembly using the Bowtie [26] short read alignment program considering the 22 chromosome assemblies.
Nimblegen probes and transcripts from ENSEMBL (version 46) and RefSeq (UCSC MM8) were mapped based on coordinates of the mouse genome assembly MM8.
Mouse miR-147 had not been deposited in miRBase because it did not have a sufficiently good match to the mouse genome assembly.
To maintain consistency between mouse genome assembly and our alignment coordinates, the multi-species alignment was compressed such that the mouse sequence is ungapped.
In the current mouse genome assembly (NCBI Build 36), the clone has five different matches with 19 identities each, but still no better match.
We downloaded Perlegen's mouse SNP genotype data [50] (release 4, where SNP coordinates are given relative to Build 37 of the mouse genome assembly, mm9) from http://mouse.perlegen.com.perlegen.com
