This repo contains functions for simulating from mechanistically-informed models of somatic hypermutation. The package can be installed using
pip install .
The primary function is simulate_sequences_abc
, so for example,
from SHMModels.simulate_mutations import simulate_sequences_abc
simulate_sequences_abc("/Users/julia/GitHub/shmr/inst/extdata/gpt.fasta",
"data/aid_logistic_3mer.csv",
context_model_length = 3,
context_model_pos_mutating = 2,
n_seqs = 1,
n_mutation_rounds = 3,
ss_file = "for_nnet_ss.csv",
param_file = "for_nnet_params.csv",
sequence_file = "for_nnet_sequences.csv",
n_sims = 100000,
write_ss=False,
write_sequences=True)
will simulate sequences from a 3-mer model, second position mutating, one sequence per parameter setting, with three rounds of mutation.
The sequences will be written to the file for_nnet_sequences.csv
and the parameters used to generate those sequences will written to for_nnet_params.csv
.
Summary statistics are a bit deprecated at this point.