Branching Evolutionary Processes Simulation

Cancer evolution simulation identifies possible principles underlying intratumor heterogeneity

Introduction
Simulation Explorer
Download

Settings

Animations

On

Off

Scale

Absolute

Relative

Parameter-centered View
Statistic-centered View
Help

s

0.01

0.1

1

r

0.0001

0.001

0.01

d

4

1
2
3
4
5
6
7
8
9
10

f

0.8

0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5

ε: Population Entropy

μ: Founder Mutation Count

ρ: Average Mutation Count

λ: Population Fitness

τ: Growth Time

θ: Self-Similarity

Simulation Instances: (click for more details)

Sorry, the experiment did not yield any results.

Details

s:
r:
d:
f:

Population Entropy (ε):
Founder Mutation Count (μ):
Average Mutation Count (ρ):
Population Fitness (λ):
Growth Time (τ):
Self-similarity (θ):

Mutation profile of each cell in a simulated tumor. 500 cells were sampled from the simulated tumor and their mutation profiles were shown as a clustered heatmap. Row and columns represent gene and cells, respectively. left blue bars indicate driver genes and top colored indicates each clone.

The simulated tumor. Each clone was colored according to top colored bar of the mutation profile heatmap.

A principal component plot of cells. Principal component analysis was applied to the mutation profiles and each cell was plotted based on loadings of the top three principal components. Note that colors of each clone was obtained from the three dimensional coordinates.

Self similality of the cell-wise dendrogram. From the cell-wise dendrogram of the mutation profile heatmap, we obtained b clusters by cutting the dendrogram at the height a. the relationship between a and b was plotted in log scale. vertical and horizontal axes denote b and a, respectively. Since their linear relationship indicates self similality, we performed linear regression between log(b) and log(a), and assumed that the dendrogram has self-similarity if R² > 0.95

Genetic distance matrix among cells. Distance between each pair of cells was calculated from the mutation profiles and presented as a heatmap.

A histogram of the mutant cell frequencies. For each gene, a fraction of mutated cells was calculated and the distribution was presented as the histogram.

Help

Here, simulation results are interactively explored. Simulations were performed for different combinations of 4 parameter values: s (stem cell hierarchy), r (mutation rate), d (number of driver genes) and f (strength of driver mutations). From 20 independent simulation instances, averaged 6 statistics were obtained: ε (population entropy), μ (founder mutation count), ρ (average mutation count), λ (population fitness), τ (growth time) and θ (self-similarity). In Parameter-centered View, d-f heatmaps of the 6 statistics are displayed in one page for selected s and r values. In Statistic-centered View, d-f heatmaps of each statistics for all examined combinations of s and r values are displayed in one page. Colors of violin plots at the right of each d-f heat map indicates scales of the heatmap. If you use absolute scale, the scales of d-f heatmaps of each statistic are common among all examined combinations of s and r values. If you use relative scale, the scale is adjusted for each d-f heatmap. If you click a cell in d-f heatmaps, images of 5 simulation instances from the corresponding parameter setting will appear below d-f heatmaps. By clicking each image, you can get details of the simulation instance.