Novoalign is suitable for aligning 454 paired end reads, just include -i option and set fragment orientation to ++ and use appropriate vales for frgment length and standard deviation.
It may be appropriate to decrease gap open penalty to improve alignment of reads with homopolymer run length errors. We have not had the chance to experiment with this.
This was tested using a set of E.Coli reads, SRR001355, from the NCBI Short read Archive.
novoalign -d ecolik12.nix -f SRR001355_1.fastq.gz SRR001355_2.fastq.gz -i ++ 3300 1000
Run on a dual core AMD PC with 8Gbyte RAM.
# novoalign (2.07.00MT – Aug 5 2010 @ 18:45:42) – A short read aligner with qualities.
# (C) 2008 NovoCraft # Licensed to Novocraft Internal Use # novoalign -d /wd5/db/ecolik12.nix -f SRR001355_1.fastq.gz SRR001355_2.fastq.gz -i ++ 3300 1000 # Interpreting input files as Sanger FASTQ. # Index Build Version: 2.7 # Hash length: 9 # Step size: 1 # Paired Reads: 62654 # Pairs Aligned: 55810 # Read Sequences: 125308 # Aligned: 124027 # Unique Alignment: 123305 # Gapped Alignment: 13026 # Quality Filter: 0 # Homopolymer Filter: 0 # Elapsed Time: 85.824 (sec.) # CPU Time: 2.7 (min.) |