Basic modified by Martin S. Lindner

Martin S. Lindner — Sat, 26 Sep 2015 14:19:31 -0000

Basic usage and examples

Similar to conventional read mappers, HiLive is a command line tool and is configured by command line parameters. It consists of two parts:

hilive, the mapper itself, and
hilive-build, the tool to prepare the reference genome k-mer index.

The first step is to create the k-mer index from the reference genome. The runtime, memory consuption, and size of the index scale exponentially with k and linearly with the size of the reference genome. Note that the k of the hilive-build tool (and thus the k-mer index) matches the hilive mapper. See [Requirements] to learn how HiLive is built with a different k.

In default mode, HiLive requires only three parameters: the location of the BCL files, the location of the k-mer index, and the number of cycles (aka read length). The number of cycles can also be set below the real read length to obtain the alignments for the inclomplete reads already before the sequencing run is finished.

hilive-build

hilive-build requires the reference genomes in a single FASTA formated file. The file may contain multiple records, such as chromosomes, organisms, etc. (multi-fasta format).

hilive-build input.fa

The index is written to the file input.fa.kix.

If the index is built from a large or highly repretitive reference genome, it makes sense to use trimming, i.e. removing k-mers from the index that occurr more than x times in the index. This can strongly reduce run time and is highly recommended. The following example discards k-mers that occur more than 1000 times:

hilive-build -t 1000 -o trimmed.kix input.fa

The index is written into the file trimmed.kix

hilive

To map reads from a 100bp sequencing run using default settings call

hilive /path/to/BaseCalls /path/to/index.kix 100

The BaseCalls directory is located in <run directory>/Data/Intensities/BaseCalls, where the <run directory> points to the location of the run directory on the HiSeq or MiSeq machine (e.g. a Samba share). Ideally you prepare and start the sequencing run and then call HiLive. Then, HiLive will wait for the first BCL files to be written by the sequencer. There is an automatic time-out after one hour.

We highly recommend to adjust the numbers of threads used by HiLive with -n. If possible, make use of all cores on the machine!

[Home]

Recent changes to Basic

Basic modified by Martin S. Lindner

Basic usage and examples

hilive-build

hilive