Genesis 2: Software for visualisation of Admixture and PCA charts.

Machanick, P.*1, Buchmann, R.2,3, Wandera, H.4, Panji, S.5, Sengupta, D.4, Hazelhurst, S.4,6

1 Department of Computer Science, Rhodes University, Makhanda, South Africa
2 School of Computer Science, University of the Witwatersrand, Johannesburg, South Africa
3 Current affiliation Discovery Health, Johannesburg, South Africa
4 Sydney Brenner Institute for Molecular Biocience, University of the
Witwatersrand, Johannesburg, South Africa
5 CBIO, University of Cape Town, South Africa
6 School of Electrical & Information Engineering, University of the Witwatersrand, Johannesburg, South Africa

Structure and Principal component analysis (PCA) plots are important tools for the analysis and presentation of data reporting population structure and genome-wide association study experiments. Tools such as eigen and admix take the genotype data and produce the structure or PCA data which then has to be visualised. Creating good quality diagrams can be a challenge because aesthetic decisions that profoundly affect the understandability and usability of the diagrams are highly data dependant, and can be very time-consuming. For example, what colours should be used for which populations, the font size of labels, the order of populations, the need to have consistency between diagrams and even label and font sizes are all affected by the number of populations and the relationships between them, the number of participants, and the clustering or admixture found. Conceptually this is a very simple task – however, it is tedious, especially for the non-specialist. Genesis is an interactive program for producing high quality structure and PCA plots for screen and paper. It uses the results of standard tools such as admix, clumpp and eigen as input. The user is able to interactively produce plots, controlling which populations and individuals should be used and their order, the colours and order of colours. For PCA plots can be produced. For structure charts, multiple charts can be placed in one picture. Some wrapper scripts are provided for converting from output of other common tools. The Genesis tool is an open-source tool running on Windows, Linux and MacOS X. It allows the rapid and interactive production of principal component and structure charts.