Correct doubling time for simulations where new genes are present but knocked out
Previously, we were initializing the RNA sequence read counts for new genes to be a small, non-zero value. Even though this did not lead to any expression of new genes, the small changes to the RNA synthesis probabilities of other genes led to a slightly longer doubling time.
To resolve this and more accurately model the impact of new genes on the existing chromosome, we decided to change the new genes to always have 0 RNA sequence read counts, so that we could be sure they did not have unintended side effects in the ParCa.
This change meant that the method we used to modify the final expression of new genes using variants also needed to be updated. The new method loads in small nonzero baseline expression parameters (which are the same for all new genes) before applying the multiplicative factor changes.
Merge new gene variants into a single variant
During the development process, we created separate variants for modifying new gene expression, modifying new gene expression and translation efficiency, and inducing/knocking out modifications to new gene expression and translation efficiency at a specific generation.
Because the induction simulations have proven to be more representative and less dependent on initialization effects, we intend to use only the features of the latter variant of the list above moving forward.
I added the ability to specify a media condition via the variant index to models/ecoli/sim/variants/new_gene_expression_and_translation_efficiency_internal_shift.py to create the merged (and now only) new gene variant, models/ecoli/sim/variants/new_gene_internal_shift.py.
Active RNAP comparison plot
Similar to models/ecoli/analysis/comparison/active_ribosome_counts_histogram.py, but for active RNAP counts.
Sample output for a wildtype simulation in minimal media (reference) and rich media (input) is shown below (single seed, 12 gens).
active_rnap_counts_histogram.pdf
This pull request has a few main components:
models/ecoli/sim/variants/new_gene_expression_and_translation_efficiency_internal_shift.pyto create the merged (and now only) new gene variant,models/ecoli/sim/variants/new_gene_internal_shift.py.models/ecoli/analysis/comparison/active_ribosome_counts_histogram.py, but for active RNAP counts.