gustavoem/signetms

SigNetMS stands for Signaling Network Model Selection

signetms's Issues

Fix jump distribution

Currently, our lognormal jump is buggy since we can't actually control it's mean and variance, which is primordial for MCMC jumping distributions.

Adaptive proposal distribution on MCMC

Pre-requisite: issue #9

Create posterior distribution plot

This looks like a good python library for this: https://seaborn.pydata.org/tutorial/distributions.html

Rewrite first phase of BIBm sampling

RandomParameters set in xml file.

We should be able to define the prior of a parameter inside the xml model definition.

Create pipeline to automate testing

Use Metropolis-Hastings algorithm instead of Metropolis

According to Supplementary Materials, proposal distribution should be log-normal. This implies that our jumping distributions are not symmetric, since that when jumping from theta = x the mean of the proposal should be x.

When proposal distributions are not symmetric, we should use the Metropolis-Hastings algorithm for sampling, instead of Metropolis algorithm.

How to determine observation scale?

Fix numerical integration of ODES

ODEintWarning: Excess work done on this call (perhaps wrong Dfun type)

Start of theta chains

Should we make different samples to each chain starting values?

Set master to use populational MCMC correctly

Work with log-likelihoods

"To avoid computational overflows and underflows, one should compute with the logarithms
of posterior densities whenever possible". (Bayesian Data Analysis)

Fix exploding jumps on MCMCInitialization

Currently, we use a lognormal distribution to sample the jump from one theta to another. That's causing an overflow because we consider that the jump is X_0 - mean (X) where X_0 is a value sampled and X_0 is the mean of the distribution of X_0. If X_0 is lognormal, than 'mean (X)' can easily become a very large number.

Define experimental measuremente as a mathematical expression.

Currently, you can only evaluate likelihood of the state of a variable of the system. Alternatively, the user may want to calculate an expression, such as ERKPP / ERK instead of just ERK.

Improve README

Create API documentation

Maybe use some tool to automatically generate the documentation.

Fix compartment parsing

We currently have a very non-specific treatment for compartments, removing them from the formulas when they have specific names.

Fix target function on MCMC

We are currently using p (y | theta) as a target function. Actually, our target function is p (theta | y).

Even though we can't calculate p (theta | y), we are able to calculate the ratio p (theta_A | y) / p (theta_B | y).

Fix observation error.

Observation error, sigma, is a parameter that should also be estimated on theta chains.

Prior probabilities are zero

On some models the prior probability is zero. We should calculate the log of the probability instead.

Read xml experiment files

Fix proposal distribution

We should clear parallel_map when execution stops.

Sometimes the marginal likelihood can fail with an exception and leave a ProcessPool opened in parallel_map.py.

On commit ed54dc3 we provided a quick and dirty fix.

Check if priors defined for all model parameters

Fix ODE solution when experiment doesn't start on time zero

The current implementation erroneously assume that the time points for solutions start on time zero.

Understand compartments

On model goodwin3.xml there's a compartment in every reaction rate. I'm not sure of what that means.

Figure out observation error distribution

According to BIBm paper, experiment error can be Gamma (2.0, 3333.0) distributed, but we are not sure if it should be sampled only once or every time we calculate a likelihood.

Allow verbose (and not verbose) run

Receive experiments as argument

Fix param prior definition and create tests

Param priors do not have correct names!

Create requirements.txt

Define experimental error in priors file

Allow SBML to create and remove interactions

Fix docstrings

Docstrings are not realy following any pattern

Describe and add cluster run to README

Generalize MetropolisHastings and write tests

Investigate jump distribution on AdaptiveMCMC

Only allow integration after system has been initialized

On ODEs, there should exist a method that prepares the system, creating the sympy autowrap for the system and jacobian. System evaluation should only be possible after using this initialization method.

Adapt proposal variance

According to supplementary materials: "The proposal distribution employed in the initialization stage is Normal on a logarithmic scale, and we were adapting the variance of the proposal distribution according to a local acceptance rate measured on each 1,000 of steps, keeping this rate between 0.25 and 0.4".

Test trapezoidal rule to solve the thermodynamic (power-posteriors) integral.

Girolami might have used this estimator instead of the estimator used on BIBm.

Fix Michaelis constants distribution

According to BIBm paper, Michaelis constants should have a Gamma (2.0, 3333.0) prior.

Correctly determine small pdf points in MultivariateLognormal

Multiple testing scripts

We have created a few statistics test that are very time consuming to be ran every time. It may be more suitable to have a small test set that does not contain many statistical tests.

Fix jump proposals

The current jumps do not have average of zero (the chain starts to increase values and never come back).

Receive priors as argument

Estimate covariance of theta parameters.

Parse lambda definitions on SBML models.

Some models from bioinformatics define lambda functions that are used to describe the rate of reactions.

Add test coverage information

Add license

Calculate Marginal Likelihoods