1. Process raw data
Your raw data contains up to ~960,000 lines of text such as above, spelling out the 'genotype' — a combination of two letters — for each position that is read on each chromosome. We count how many times we see each combination per chromosome.
2. Assign colors
Each two-letter combination is represented by a distinct color. We also cover the cases where there was "no call", or a position could not be read because of insertions and deletions.
3. Draw shapes
We draw a shape for each genotype. The size corresponds to the count, and so does the angle at which it is drawn.
4. Stack and repeat
Then the shapes are overlaid: the largest at the bottom, the smaller ones on top, so we can see them all.
We repeat this for each chromosome.
