Novoalign

Paired End Reads

Paired end reads arise when a single fragment of DNA has been read from both ends, generating a pair of reads possibly separated by some sequence that was not included in the reads. The two reads are on opposite strands.

The main issues with paired end reads is short fragments such that the two reads overlap or if the fragment length is shorter than the read length then the reads extend into adapter sequence. Novoalign can detect short fragments and paired end reads that have extended into adapter and trim the adapter sequence, just add the option -a to the command line.

Mate Pairs

Illumina Mate pairs are from long fragments that are circularised and then fragmented again. A Biotin marker on the circularisation junction is used to enrich for fragments that overlap the junction.

There are several considerations for alignment of Illumina mate pairs:

1. The Biotin enrichment is not perfect and there are typically 30-50% of fragments that do not overlap the circularisation junction and hence align like paired end reads
2. It’s also likely that the circularisation junction is sufficiently close to the fragment end that a read will cross the junction. The Novoalign mate pair algorithm can detect this and split the read at the junction.
3. It’s also possible that the secondary fragmentation produces short fragments where the two reads overlap or extend into adapter.

454 Paired End

454 paired end reads are produced by a circularisation and a secondary fragmentation step. Alignment of reads is similar to ABI SOLiD in that the two reads of the pair are on the same strand.

NovoalignCS

Mate Pairs

Paired End