Figure 1.
Schema of the Computational Epigenetic Micro-model.
Structure and layers of our computational model closely represent known epigenetic mechanisms. The master Object is List-Block which generates a Block of genes. Contained inside each Block are the DNA and Histone objects forming Nucleosome unit. There are 8 histone objects - pairs of H2A,H2B,H3 and H4 along with one H1 object. Each of the histone objects are updated with the modifications over each time-step during the simulation. Hence during each time step, the model is aware of the Histone modifications and DNA Methylation which defines the system evolution.
Figure 2.
General representation of histone states in our model.
The number of modifiable amino acids chosen for each histone type differs. In general, each modification is encoded as a number - Acetylation as “1”, Methylation as “2”, Phosphorylation as “3” and no modifications as “0”. The string of numbers or the current Histone state represents the possible combination of modifications within that particular histone type.
Table 1.
Amino acid positions and modifications.
Figure 3.
Probability of shift between a histone state and its neighbour.
Only one change is possible at each time step hence each histone state can potentially shift to only one of its specific neighbours. * = Current state. # - neighboring state. Probabilities of shift ([0,1]) can be given by the user initially.
Table 2.
Compression of H3 type histone.
Figure 4.
Average Transcription Progression derived for 16 Promoters over 5000 iterations during 3 different simulation runs.
The third type of interaction between Transcription rate and DNA methylation level (or percentage) was observed here. Transcription rate (25 time-steps = 1 time-interval) is inversely proportional to DNA methylation level (decided by user in this case, for testing purposes).
Figure 5.
Evolution of H4 (H4-1 and H4-2) histone states in the 16 promoters for 10 different datasets during low DNA methylation levels (<0.15 or 15%).
H4-1 and H4-2 histone states were tested with 10 datasets of random probability values (represented by colors in the graph).
Figure 6.
Analysis of Average percentage visits of H3 histone states containing Lysine Acetylation .i.e K6 in 16 promoters after 5000 iterations (for low levels of DNA Methylation (<0.15 or 15%)).
States containing Lysine acetylation are visited the most. Hence we analyse the the average of percentage visitation of model to all other modifications (except Lysine Acetylation - K6 in Figure) during the simulation. Here each unit in the X-axis represents an amino acid–position in H3 array–number of Amino acids–Modification possible. The Y axis elaborates on the average percentage visitation of H3 states that contain the modification depicted in each unit of X axis.
Figure 7.
Analysis of Average percentage visits of H3 histone states containing Lysine Methylation i.e. K5 in 16 promoters after 5000 iterations (for high levels of DNA Methylation (<0.85 or 85%)).
States containing Lysine Methylation are visited the most. Hence we analyse the average of percentage visitation of model to all other modifications (except Lysine Methylation K5) during the simulation. Each unit in the X-axis represents an amino acid–position in H3 array–number of Amino acids changeable–Modification. The Y axis elaborates on the average percentage visitation of H3 states that contain the modification given on the X axis.
Figure 8.
Evolution of H4 (H4-1 and H4-2) histone states in the 16 promoters for 10 different datasets during high DNA methylation levels (<0.85 or 85%).
H4-1 and H4-2 histone states were tested with 10 dataset of random probability values (represented by colors in the graph).
Figure 9.
A Comparison between the average (of all 20 test results obtained for H4-1 and H4-2) preferences of H4 states for high and low DNA Methylation Levels.
Error bars represent the standard deviation calculated from the total number of visits, for every H4 histone state (occupancy) during the simulation.