# How to interpret fish plot in R

I am working on subclonal evolution of multiple myeloma. After literature survey, I found people use Fish plot to represent tumor evolution. The example of fish plot can be found here. But I am not able to interpret this plot. How this plot represent tumor evolution and subclones in tumor. Can anybody explain with example. Thanks.

• Have you read the paper cited in your link? Doesn't that explain them? We can't help you understand a plot if you don't even tell us what is being plotted, what the y-axis represents. – terdon Mar 11 '19 at 10:25
• @terdon : This is what I did as first step. But in their paper, they have not given any such explanation or interpretation of Fish plot (implementation info given only) . Paper can be found at (bmcgenomics.biomedcentral.com/articles/10.1186/…). I was hoping to get some useful information in paper but not. – Lot_to_learn Mar 11 '19 at 10:32
• The population (cells) is going through a bottleneck, whether that is a population size, aneuploidy, dysregulation I've no idea, the y-axis is not labelled. If it's not an oncovirus I'm stumped. Oh it's longitudinal study BTW (obviously) – M__ Mar 11 '19 at 12:00

Each color is a clone (set of mutations) or an individual mutation (depends on how the input data was generated). This is a cancer with the bottleneck likely being induced by a treatment (e.g. UV, chemo, drugs, etc...)

Based on the vignette this is likely an AML (leukemia).

It looks like grey are healthy cells. At day 0 a specific clone (red) existed at a low frequency (~20-40%). The patient (or cells) was then exposed to a treatment and on day 150 another measurement was taken. In this case two novel clones (subclones of red) have taken over the population. They seem to co-exist or are co-dominate clones with orange being 50-60% of the population and yellow being 30-40%.

The example says "Simple plot with multiple subclones" so: the orange and yellow populations have additional (likely mutually exclusive) mutations that were not present in the red clone - but biologically also have all the mutations in the red clone. So my first assumption was right - each color is a proxy for a set of mutations.

You could have generated the mutation data (i.e. allele frequencies) from a wide variety of sources (scRNA-Seq, exome-sequencing, PCR genotyping, etc...)

• Thanks for this simplistic explanation. I want to know what does that narrow gray color line, width of each sub-region represent in the above figure (between day 0 and day 150). In the same example (https://www.rdocumentation.org/packages/fishplot/versions/0.4) the other graph with More complex clonal structure, in day 403, blue region continue till day 403 but width reduces and green/red region starts very late in day 403 region. Does that mean that sub-clone represented by blue region still exists in tumor on day 403 and green/red sub-clones develop on last time span? – Lot_to_learn Mar 12 '19 at 7:14
• Similarly, in figure Multiple independent clones, many timepoints, your explaination is not perfectly applicable. One sub-clone is coming separately from healthy cells. But as per your explanation, all the sub-clones should lie in between healthy cell region (even for multiple clones). – Lot_to_learn Mar 12 '19 at 7:17
• Its easy, the graph infers treatment was counter-productive. There is an initial clonal expansion of a cancerous cell (red), within healthy tissue (grey). The cancerous cells appears to be controlled by treatment, but healthy tissue is also destroyed (bottleneck). Two new clonal expansions emerge in place of the red/grey cells post-treatment and rapidly become dominant. Thus treatment has lead to a more aggressive cancer, i.e. counter-productive. In population genetics its a "selective sweep", orange/yellows clones are likely present in the originally but couldn't become dominant. – M__ Mar 13 '19 at 16:22