GARD terminates with error

[skyungyong]

Dear developers,

I am running GARD on alignment files, and GARD terminates with error messages for some of the alignments.

Here is the some part of the error message. How would this be happening?

Thank you for your help!

• Analysis Version: 0.1

type: nucleotide rv: GDD

How many site rate classes to use (permissible range = [2,10], default value = 4, integer): rate-classes: 3 Loaded a nucleotide multiple sequence alignment with 13 sequences, 4119 sites (330 of which are variable) from OG013.cds.msa.fa Minimum size of a partition is set to be 23 sites

Fitting the baseline (single-partition; no breakpoints) model

• Log(L) = -8464.43, AIC-c = 17001.52 (36 estimated parameters)

Performing an exhaustive single breakpoint analysis

Best sinlge break point location: 1029 c-AIC = 16599.34261043844

Performing multi breakpoint analysis using a genetic algorithm

Best break point locations: 1386, 2874 c-AIC = 16389.29919469103

Done with 3 breakpoint analysis. Best break point locations: 982, 1513, 2874 c-AIC = 16313.70766359857

Done with 4 breakpoint analysis. Best break point locations: 812, 1434, 3097, 3556 c-AIC = 16228.65249652059

Done with 5 breakpoint analysis. Best break point locations: 123, 812, 1434, 3097, 3556 c-AIC = 16194.05418586989

Best break point locations: 123, 812, 1436, 2874, 3313, 3655
c-AIC = 16155.52421103128

Error:

Master node received an error:Internal error in GARD -- c-AIC INCREASED between two consecutive generations in call to assert(gard.previousBest_cAIC>=gard.currentBest_cAIC, error_msg);

Function call stack 1 : assert(gard.previousBest_cAIC>=gard.currentBest_cAIC, error_msg);

---

2 : ExecuteCommands("assert (statement, error_msg)", /usr/local/lib/hyphy/TemplateBatchFiles/libv3/);

3 : [namespace = gard] io.CheckAssertion("gard.previousBest_cAIC >= gard.currentBest_cAIC","Internal error in GARD -- c-AIC INCREASED between two consecutive generations");

4 : namespace

Step 0.populationSize=32;

Step 1.Branch under condition 'populationSize<mpi.NodeCount()-1' to populationSize=mpi.NodeCount()+1; else mutationRate=0.8;

Step 2.populationSize=mpi.NodeCount()+1;

Step 3.mutationRate=0.8;

Step 4.rateOfMutationsTharAreSmallShifts=0.5;

Step 5.maxFailedAttemptsToMakeNewModel=7;

Step 6.cAIC_diversityThreshold=0.001;

Step 7.cAIC_improvementThreshold=0.01;

Step 8.maxGenerationsAllowedWithNoNewModelsAdded=5;

Step 9.maxGenerationsAllowedAtStagnant_cAIC=100;

Step 10.addingBreakPointsImproves_cAIC=TRUE;

Step 11.numberOfBreakPointsBeingEvaluated=1;

Step 12.Branch under condition 'addingBreakPointsImproves_cAIC' to numberOfBreakPointsBeingEvaluated+=1; else

Step 13.numberOfBreakPointsBeingEvaluated+=1; Step 14.generationsAtCurrentBest_cAIC=0; Step 15.generationsNoNewModelsAdded=0; Step 16.parentModels=gard.GA.initializeModels(numberOfBreakPointsBeingEvaluated,populationSize,numberOfPotentialBreakPoints,bestOverallModelSoFar); Step 17.terminationCondition=FALSE; Step 18.generation=0; Step 19.Branch under condition 'terminationCondition==FALSE' to childModels=gard.GA.recombineModels(parentModels,populationSize); else io.ClearProgressBar(); Step 20.childModels=gard.GA.recombineModels(parentModels,populationSize); Step 21.interGenerationalModels=parentModels; Step 22.interGenerationalModels*childModels; Step 23.gard.GA.evaluateModels(interGenerationalModels); Step 24.selectedModels=gard.GA.selectModels(interGenerationalModels,populationSize); Step 25.Branch under condition 'gard.GA.modelSetsAreTheSame(selectedModels,parentModels)' to generationsNoNewModelsAdded+=1; else generationsNoNewModelsAdded=0;

[spond]

Dear @skyungyong,

Thank you for reporting this issue. I will issue a fix shortly and post an update here.

Best, Sergei

[spond]

Dear @skyungyong,

I updated the 2.5.4 release to correct this error which was caused by an oversight on my part. Please update your HyPhy distribution and try again.

Thanks, Sergei

[skyungyong]

Dear @spond,

Thank you for the quick update! GARD works now perfectly.

best wishes.