Release Notices

Hi,

We've just posted a new release of Novoalign to the download page. The changes are fairly minor.

We've run some ION Torrent alignments with good results, and just for interest we added a module to collect insert/deletion statistics for reads. This can be used to chart homopolymer run length errors as a function of run length and position in the read. There's some details in our User Guide. The indel counts can be collected for any reads and we've looked at ION Torrent, 454 & MiSeq reads. We expect ION Torrent results to keep improving.

Full change list is below.

Enjoy,
Novocraft

Novoalign

1. Fix Seg Fault that occurred if -Q option was used with single end reads.
2. Add option to collect statistics on homopolymer run length errors.
   --hpstats [filename]
   These can be charted using included R script IONTorrent.R
   IONTorrent.R -f indelstats.tsv -r indelcharts.pdf
   The data collection and script will work for 454 Paired & Illumina reads.
3. Add option to check that paired end read files are in the same order. This involves checking that the headers of the two reads in a pair match within a specified Hamming Distance.
   --hdrhd [9|off] Controls checking of identity between headers in paired end reads. Sets the Hamming Distance or disables the check. Default is a Hamming Distance of not more than 1. Processing will stop with appropriate error messages if Hamming Distance exceeds the limit.
   The default of 1 should be valid for standard Illumina headers.
4. In Paired end mode, previously reads shorter than log4(reference length) would not be aligned, we now honour the -l setting, allowing reads as short as the index k+s-4.

NovoalignMPI

1. Fix: If the number of reads input to an MPI run was insufficient to transfer at least one buffer of reads to each slave then the MPI process would hang after processing all reads. Buffers are 16Kbyte.
2. Compile and link using MPICH2 V1.3.2p1

Novoindex

1. Added an option to control the number of threads used for indexing. Option is -t 9

Novobarcode

1. Add option (--QSEQ_OUT) to force qseq format output if input is qseq and index tags are embedded in reads. If not used then reads are converted to fastq format.
2. Add option -d folder that sets base folder name for demux'd read files. The folder should exist.