nicfel/mascot

BEAST2 implementation of the Masco approximation of the structured coalescent with forwards/backwards algorithm for node state calculation

MASCOT

MASCOT is a BEAST2 package that allows to use MASCOT with other BEAST2 utilities. A tutorial on how to use MASCOT can be found here

License

The content of this project itself is licensed under the Creative Commons Attribution 3.0 license, and the java source code of esco is licensed under the GNU General Public License.

mascot's Issues

Mascot not compatible with bModelTest

Currently, bModelTest does not show up in BEAUti when selecting the Mascot template. Adding this line

    <mergepoint id='aux-sitemodel-panels'/>

to the templates will fix this.

GLM constant error dimension initialization error

The dimension of the error term in the GLM model is not initialized in the loglinear class

Potential math error in exponential coalescent

Hi Dr. Müller,

Really sorry if I'm mistaken here, but I think there might be an error in the math in the getNeTime function in mascot.parameterdynamics.ExponentialNe, although it's possible I'm misunderstanding how the parameters are defined.

If NeNull is the population size at the most recent calendar time and getNeTime is supposed to return the population size at time t in the past then I think it should be NeNull * Math.exp(-t*growthRate).

E.g. with a growth rate of 0.25 and a NeNull of 100 you would expect the population size 10 years in the past to be 100*exp(-0.25 * 10) = 8.2.

This is how (what I presume is) the analogous function in the regular exponential coalescent function is written.

Again, if I'm misunderstanding I apologize, but I can't wrap my head around the function as written.

Mascot/src/mascot/parameterdynamics/ExponentialNe.java

Line 34 in 43100a5

public double getNeTime(double t) {

Mascot and partition bug

To replicate:

• Import 2 partitions into Mascot template, so now there are 2 trees and 2 structuredCoalescent objects.
• link trees in partition panel

There are still Ne and migration parameters for the second partition in the state, that have a prior, and operators on them. They do no harm, but are confusing and slowing things down a bit.

Remove Jblas dependency in future release

The jblas dependency can cause errors on some machines. As non of its computing machinery is used anyway, it should be removed in a future release

The output of the exponential coalescent should be logNeNull instead of NeNull

The output of the log file for the exponential coalescent should make it clear that it's log and not real space

Mascot does not appear in the drop down menu at the Priors panel

I am using Beast v2.6.2, Java v1.8.0_212, Mascot v2.0.1 in a Windows 10 Pro system. Thank you for your contribution.

Add error reporting when adding predictor

When adding a predictor that has some error in it, no error is reported at the moment, making it hard to troubleshoot