Back

Introducing and benchmarking the accuracy of cayenne: a Python package for stochastic simulations

Kishore, D.; Chandrasekaran, S.

2020-10-10 systems biology
10.1101/2020.10.10.334623 bioRxiv
Show abstract

Biological systems are intrinsically noisy and this noise may determine the qualitative outcome of the system. In the absence of analytical solutions to mathematical models incorporating noise, stochastic simulation algorithms are useful to explore the possible trajectories of these systems. Algorithms used for such stochastic simulations include the Gillespie algorithm and its approximations. In this study we introduce cayenne, an easy to use Python package containing accurate and fast implementations of the Gillespie algorithm (direct method), the tau-leaping algorithm and a tau-adaptive algorithm. We compare the accuracy of cayenne with other stochastic simulation libraries (BioSimulator.jl, GillespieSSA and Tellurium) and find that cayenne offers the best trade-off between accuracy and speed. Additionally, we highlight the importance of performing accuracy tests for stochastic simulation libraries, and hope that it becomes standard practice when developing the same. The cayenne package can be found at https://github.com/Heuro-labs/cayenne while the bench-marks can be found at https://github.com/Heuro-labs/cayenne-benchmarks

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

1
PLOS Computational Biology
● 1863 papers in training set
Top 0.4%
32.8%
Similar papers in this journal
2
BMC Bioinformatics
○ 457 papers in training set
Top 0.1%
19.5%
Similar papers in this journal
50% of probability mass above
3
Bioinformatics
◐ 1204 papers in training set
Top 2%
10.3%
Similar papers in this journal
4
PLOS ONE
● 5266 papers in training set
Top 30%
4.6%
Similar papers in this journal
5
Frontiers in Systems Biology
○ 10 papers in training set
Top 0.1%
1.8%
Similar papers in this journal
6
Journal of Theoretical Biology
○ 162 papers in training set
Top 1%
1.8%
7
Scientific Reports
○ 3612 papers in training set
Top 50%
1.8%
8
Ecology and Evolution
○ 267 papers in training set
Top 3%
1.8%
9
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 4%
1.6%
10
Bioinformatics Advances
◐ 203 papers in training set
Top 3%
1.4%
11
Neuroinformatics
○ 46 papers in training set
Top 0.6%
1.2%
12
iScience
○ 1154 papers in training set
Top 29%
1.1%
13
Ecological Modelling
○ 28 papers in training set
Top 0.4%
1.1%
14
Methods in Ecology and Evolution
○ 176 papers in training set
Top 1%
1.1%
15
Bulletin of Mathematical Biology
○ 92 papers in training set
Top 1%
1.0%
16
Gigabyte
● 62 papers in training set
Top 1%
0.9%
17
Briefings in Bioinformatics
◐ 354 papers in training set
Top 7%
0.9%
18
npj Systems Biology and Applications
○ 125 papers in training set
Top 2%
0.9%
19
PeerJ
◐ 308 papers in training set
Top 10%
0.9%
20
Computer Methods and Programs in Biomedicine
○ 28 papers in training set
Top 1%
0.6%
21
in silico Plants
◐ 27 papers in training set
Top 0.3%
0.5%
22
Royal Society Open Science
● 214 papers in training set
Top 8%
0.5%
23
IFAC-PapersOnLine
○ 13 papers in training set
Top 0.3%
0.5%
24
IEEE/ACM Transactions on Computational Biology and Bioinformatics
● 38 papers in training set
Top 1%
0.5%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.