Often, researchers wish to compare groups of isolates against each other- for example, compare samples from one geographic region with those from another; or compare drug-resistant with drug-sensitive isolates. MapSeq allows users to define arbitrary groups of samples, and then visualize the genotype calls for the two groups in any region of interest, highlighting differences in the distribution of alleles in the two groups. These differences help users to identify significant mutations in one group or the other, which may be related to epidemiological phenomena.

Please note that you will need viewing privileges to see the genotyping data for a number of samples. These privileges are usually granted by the investigator who submitted the samples (see more details here). However, a number of samples have publicly-viewable data, which you can freely browse.

The figure below shows an example of a comparative data analysis. Here, we have formed two groups: one with samples from South-East Asia (SEA, x samples) and one with samples from Papua New Guines (PNG, x samples). The figure show the distribution of genotyped alleles in chromosome 7, starting from position 300,000. At each position, MapSeq measures mutual information (MI), a value that measures how different the two allele distributions are, whose strength is shown by the background colour of the central column. (deep red denoting positions where MI is highest). The flanking columns show the sample count for each of the alleles A, C, G and T for each group, while the last column shows annotations for the position.

The positions with the strongest MI signal (lower part of the figure) are mutations within the PfCRT gene, and its neighbouring genes. The reason for such a cluster of highly differentiated mutations in the two groups is the emergence of chloroquine (CQ) resistance in the two regions. Selection for drug-resistant mutant alleles in PfCRT caused these alleles to sweep across South-East

Asia and New Guinea; since drug resistance emerged independently in the two regions, a comparison of these regions reveals different mutations in each set- which explains why so many positions have highly differentiated alleles. As can be clearly seen, the drug-resistance "signal" extends to several neighbouring mutations, which do not play a role in CQ resistance. These mutations spread throughout the regions because their proximity to the core mutations prevented them from being "lost" by recombination. Over time, we believe that the high-MI region will shrink around the PfCRT gene.

By using this functionality, it is possible to identify region of the genome that contain mutations associated with phenotypes, or with geographical or temporal location. If you wish to know whether the call is well-supported, simply click on the sample counts, and a popup will display the number of reads for each sample in the group. If you want to further inspect the reads assembly, you can click on "View Read Pileup" which will invoke MapSeq to show the deep sequencing reads].

Why not try this? Please follow the instructions in our simple tutorial.