Help

Sequences

Submitted files should be nucleotide sequence in multi-fasta format (uncompressed). The default setup will only look at the first 1000 sequences in each file. The total size of encoded data must be below 50MB.

The first sequence file is mandatory but the second file is optional. If you have more than two sequence groups then you will need to use the command line version of compter to process these.

Background

Kmer values from compter are generally expressed relative to a background Kmer distribution. You have a number of choices how to set this and these may affect the perceived enrichment or depletion of the kmers in your data.

1. You can choose to have no background. In this case the values reported by the tool are the frequencies with which each kmer is observed. If you are doing this then we would recommend performing a fixed length kmer analysis otherwise there will be systematic differences in the observed frequencies of kmers of different lengths
2. You can construct a theoretical background based on a specific GC content. This will generally reflect the expected level of each kmer but will not show and organism specific biases (eg the general under-representation of CG in higher eukaryotes)
3. You can use a pre-calculated background. You will see a list of pre-calculated frequences for the species which have been installed on your compter installation. These represent the true observed kmer frequencies in the genome you select

If you use a background then the values reported will be log2(observed/expected) for your sample against the background.

In the command line version of compter you can calculate and save a custom background from an additional fasta file. If you have done this you can ask the admin of your compter web system to add the cmp file generated to the config directory so that it shows up as an option for future web searches.

Kmer size range

The kmer size range simply defines the maximum and minimum size of the kmer subsequences which will be counted and analysed in your data.

Due to the increasing number of permutations and loss of idependence we don't allow kmer sizes above 3 to be analysed.

Clustering

This option sets whether the ordering of the sequences you supply is changed in the heatmap to place sequences which have similar composition next to each other. If you have submitted two groups of sequences then these will be allowed to mix so you can see how well the composition separates the groups you defined. If you turn this option off then the sequences from each file will be kept together, and the ordering of sequences within each file will also be preserved.