msasim 25.10.15__tar.gz → 25.11.2__tar.gz

{msasim-25.10.15/msasim.egg-info → msasim-25.11.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: msasim
-Version: 25.10.15
+Version: 25.11.2
 Summary: A fast MSA simulator
 Home-page: https://github.com/elyawy/Sailfish-backend
 Author: Elya Wygoda

{msasim-25.10.15 → msasim-25.11.2}/_Sailfish/__init__.pyi

@@ -376,6 +376,8 @@ class modelFactory:
         ...
     def set_invariant_sites_proportion(self, arg0: float) -> None:
         ...
+    def set_site_rate_correlation(self, arg0: float) -> None:
+        ...
     def set_model_parameters(self, arg0: list[float]) -> None:
         ...
     def set_replacement_model(self, arg0: modelCode) -> None:

{msasim-25.10.15 → msasim-25.11.2}/msasim/sailfish.py

@@ -25,14 +25,14 @@ class Distribution:
                 raise ValueError(f"Each value of the probabilities should be between 0 to 1. Received a value of {x}")
         self._dist = _Sailfish.DiscreteDistribution(dist)
     
-    def draw_sample(self) -> int:
-        return self._dist.draw_sample()
+    # def draw_sample(self) -> int:
+    #     return self._dist.draw_sample()
     
-    def set_seed(self, seed: int) -> None:
-        return self._dist.set_seed(seed)
+    # def set_seed(self, seed: int) -> None:
+    #     return self._dist.set_seed(seed)
     
-    def get_table(self) -> List:
-        return self._dist.get_table()
+    # def get_table(self) -> List:
+    #     return self._dist.get_table()
     
     def _get_Sailfish_dist(self) -> _Sailfish.DiscreteDistribution:
         return self._dist
@@ -415,7 +415,10 @@ class Simulator:
         # verify sim_protocol
         if self._verify_sim_protocol(simProtocol):
             self._simProtocol = simProtocol
-            self._simulator = _Sailfish.Simulator(self._simProtocol._sim)
+            if simulation_type == SIMULATION_TYPE.PROTEIN:
+                self._simulator = _Sailfish.AminoSimulator(self._simProtocol._sim)
+            else:
+                self._simulator = _Sailfish.NucleotideSimulator(self._simProtocol._sim)
         else:
             raise ValueError(f"failed to verify simProtocol")
         
@@ -475,7 +478,8 @@ class Simulator:
             model_parameters: List = None,
             gamma_parameters_alpha : float = 1.0,
             gamma_parameters_categories: int = 1,
-            invariant_sites_proportion: float = 0.0
+            invariant_sites_proportion: float = 0.0,
+            site_rate_correlation: float = 0.0,
         ) -> None:
         if not model:
             raise ValueError(f"please provide a substitution model from the the following list: {_Sailfish.modelCode}")
@@ -503,6 +507,8 @@ class Simulator:
 
         self._model_factory.set_gamma_parameters(gamma_parameters_alpha, gamma_parameters_categories)
         self._model_factory.set_invariant_sites_proportion(invariant_sites_proportion)
+        self._model_factory.set_site_rate_correlation(site_rate_correlation)
+
         self._simulator.init_substitution_sim(self._model_factory)
 
         self._is_sub_model_init = True
@@ -556,13 +562,11 @@ class Simulator:
             msa = Msa(blocktree._get_Sailfish_blocks(),
                         self._simProtocol._get_root(),
                         self.get_sequences_to_save())
+            self._simulator.set_aligned_sequence_map(msa._msa)
 
         # sim.init_substitution_sim(mFac)
         if self._simulation_type != SIMULATION_TYPE.NOSUBS:
-            with tempfile.TemporaryDirectory() as tmpdirname:
-                self._simulator.gen_substitutions_to_dir(msa.get_length(), tmpdirname)
-                msa._msa.set_substitutions_folder(tmpdirname)
-                msa._msa.write_msa_from_dir(str(output_file_path))
+            self._simulator.gen_substitutions_to_file(msa.get_length(), str(output_file_path))
 
 
     

{msasim-25.10.15 → msasim-25.11.2/msasim.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: msasim
-Version: 25.10.15
+Version: 25.11.2
 Summary: A fast MSA simulator
 Home-page: https://github.com/elyawy/Sailfish-backend
 Author: Elya Wygoda

{msasim-25.10.15 → msasim-25.11.2}/msasim.egg-info/SOURCES.txt

@@ -13,5 +13,4 @@ msasim.egg-info/not-zip-safe
 msasim.egg-info/requires.txt
 msasim.egg-info/top_level.txt
 src/main.cpp
-src/modelFactory.cpp
-src/rateMatrixSim.cpp
+src/modelFactory.cpp

{msasim-25.10.15 → msasim-25.11.2}/setup.py

@@ -6,7 +6,7 @@ from setuptools import setup, find_packages
 
 
 
-__version__ = "25.10.15"
+__version__ = "25.11.2"
 
 # The main interface is through Pybind11Extension.
 # * You can add cxx_std=11/14/17, and then build_ext can be removed.

{msasim-25.10.15 → msasim-25.11.2}/src/main.cpp

@@ -4,6 +4,7 @@
 #include <pybind11/stl.h>
 #include <memory>
 
+#include "../libs/pcg/pcg_random.hpp"
 #include "./Simulator.h"
 
 namespace py = pybind11;
@@ -29,6 +30,8 @@ PYBIND11_MODULE(_Sailfish, m) {
             Tree
     )pbdoc";
 
+    using SelectedRNG = pcg32_fast;
+
     py::class_<Block>(m, "Block")
         .def(py::init<size_t, size_t>());
 
@@ -43,10 +46,7 @@ PYBIND11_MODULE(_Sailfish, m) {
         .export_values();
 
     py::class_<DiscreteDistribution>(m, "DiscreteDistribution")
-        .def(py::init<std::vector<double>>())
-        .def("draw_sample", &DiscreteDistribution::drawSample, "Draw a random sample according to the given distribution")
-        .def_static("set_seed", &DiscreteDistribution::setSeed, "Set seed for the random number generator")
-        .def("get_table", &DiscreteDistribution::getTable, "Get Vose's alias table (useful for debugging)");
+        .def(py::init<std::vector<double>>());
 
     py::class_<tree>(m, "Tree")
         .def(py::init<const std::string&, bool>(), "Create Phylogenetic tree object from newick formatted file")
@@ -128,32 +128,48 @@ PYBIND11_MODULE(_Sailfish, m) {
         .def("set_model_parameters" , &modelFactory::setModelParameters)
         .def("set_gamma_parameters" , &modelFactory::setGammaParameters)
         .def("set_invariant_sites_proportion", &modelFactory::setInvariantSitesProportion)
+        .def("set_site_rate_correlation", &modelFactory::setSiteRateCorrelation)
         .def("reset", &modelFactory::resetFactory);
 
 
-    py::class_<Simulator>(m, "Simulator")
+    py::class_<Simulator<SelectedRNG, 20>>(m, "AminoSimulator")
+        .def(py::init<SimulationProtocol*>())
+        .def("reset_sim", &Simulator<SelectedRNG, 20>::resetSimulator)
+        .def("gen_indels", &Simulator<SelectedRNG, 20>::generateSimulation)
+        .def("run_sim", &Simulator<SelectedRNG, 20>::runSimulator)
+        .def("init_substitution_sim", &Simulator<SelectedRNG, 20>::initSubstitionSim)
+        .def("gen_substitutions", &Simulator<SelectedRNG, 20>::simulateSubstitutions)
+        .def("gen_substitutions_to_file", &Simulator<SelectedRNG, 20>::simulateAndWriteSubstitutions)
+        .def("set_aligned_sequence_map", &Simulator<SelectedRNG, 20>::setAlignedSequenceMap)
+        .def("save_site_rates", &Simulator<SelectedRNG, 20>::setSaveRates)
+        .def("get_site_rates", &Simulator<SelectedRNG, 20>::getSiteRates)
+        .def("save_all_nodes_sequences", &Simulator<SelectedRNG, 20>::setSaveAllNodes)
+        .def("save_root_sequence", &Simulator<SelectedRNG, 20>::setSaveRoot)
+        .def("get_saved_nodes_mask", &Simulator<SelectedRNG, 20>::getNodesSaveList);
+
+    py::class_<Simulator<SelectedRNG, 4>>(m, "NucleotideSimulator")
         .def(py::init<SimulationProtocol*>())
-        .def("reset_sim", &Simulator::resetSimulator)
-        .def("gen_indels", &Simulator::generateSimulation)
-        .def("run_sim", &Simulator::runSimulator)
-        .def("init_substitution_sim", &Simulator::initSubstitionSim)
-        .def("gen_substitutions", &Simulator::simulateSubstitutions)
-        .def("gen_substitutions_to_dir", &Simulator::simulateAndWriteSubstitutions)
-        .def("save_site_rates", &Simulator::setSaveRates)
-        .def("get_site_rates", &Simulator::getSiteRates)
-        .def("save_all_nodes_sequences", &Simulator::setSaveAllNodes)
-        .def("save_root_sequence", &Simulator::setSaveRoot)
-        .def("get_saved_nodes_mask", &Simulator::getNodesSaveList);
+        .def("reset_sim", &Simulator<SelectedRNG, 4>::resetSimulator)
+        .def("gen_indels", &Simulator<SelectedRNG, 4>::generateSimulation)
+        .def("run_sim", &Simulator<SelectedRNG, 4>::runSimulator)
+        .def("init_substitution_sim", &Simulator<SelectedRNG, 4>::initSubstitionSim)
+        .def("gen_substitutions", &Simulator<SelectedRNG, 4>::simulateSubstitutions)
+        .def("gen_substitutions_to_file", &Simulator<SelectedRNG, 4>::simulateAndWriteSubstitutions)
+        .def("set_aligned_sequence_map", &Simulator<SelectedRNG, 4>::setAlignedSequenceMap)
+        .def("save_site_rates", &Simulator<SelectedRNG, 4>::setSaveRates)
+        .def("get_site_rates", &Simulator<SelectedRNG, 4>::getSiteRates)
+        .def("save_all_nodes_sequences", &Simulator<SelectedRNG, 4>::setSaveAllNodes)
+        .def("save_root_sequence", &Simulator<SelectedRNG, 4>::setSaveRoot)
+        .def("get_saved_nodes_mask", &Simulator<SelectedRNG, 4>::getNodesSaveList);
 
 
     py::class_<MSA>(m, "Msa")
         .def(py::init<size_t, size_t, const std::vector<bool>& >())
-        .def(py::init<BlockMap, tree::TreeNode*, const std::vector<bool>& >())
+        .def(py::init<BlockMap&, tree::TreeNode*, const std::vector<bool>& >())
         .def("generate_msas", &MSA::generateMSAs)
         .def("length", &MSA::getMSAlength)
         .def("num_sequences", &MSA::getNumberOfSequences)
         .def("fill_substitutions", &MSA::fillSubstitutions)
-        .def("set_substitutions_folder", &MSA::setSubstitutionsFolder)
         .def("print_msa", &MSA::printFullMsa)
         .def("print_indels", &MSA::printIndels)
         .def("write_msa", &MSA::writeFullMsa)

{msasim-25.10.15 → msasim-25.11.2}/src/modelFactory.cpp

@@ -82,9 +82,17 @@ void modelFactory::setGammaParameters(MDOUBLE alpha, size_t numCategories) {
     _state = factoryState::COMPLETE;
 }
 
+void modelFactory::setSiteRateCorrelation(MDOUBLE correlation) {
+    if (correlation < -1.0 || correlation > 1.0) {
+        errorMsg::reportError("Rate correlation must be between -1 and 1");
+    }
+    _siteRateCorrelation = correlation;
+}
+
 void modelFactory::resetFactory() {
     _state = factoryState::ALPHABET;
     _invariantProportion = 0.0;
+    _siteRateCorrelation = 0.0;
 }
 
 

msasim-25.10.15/src/rateMatrixSim.cpp

@@ -1,278 +0,0 @@
-// $Id: simulateTree.cpp 8508 2010-08-12 15:21:04Z rubi $
-#include <stack>
-#include <unordered_map>
-#include <ostream>
-#include <sstream>
-
-
-#include "../libs/Phylolib/includes/definitions.h"
-#include "../libs/Phylolib/includes/treeUtil.h"
-#include "../libs/Phylolib/includes/talRandom.h"
-#include "../libs/Phylolib/includes/gammaDistribution.h"
-#include "../libs/Phylolib/includes/codon.h"
-
-#include "rateMatrixSim.h"
-// simulateTree::simulateTree(tree*  _inEt,
-// 						   const stochasticProcess* sp,
-// 						   const alphabet* alph) :
-// 	_et(_inEt), _sp(sp),_alph(alph),_avgSubtitutionsPerSite(0.0) {
-// 	};
-
-rateMatrixSim::rateMatrixSim(modelFactory& mFac, std::shared_ptr<std::vector<bool>> nodesToSave) : 
-	_et(mFac.getTree()), _sp(mFac.getStochasticProcess()), _alph(mFac.getAlphabet()), 
-	_invariantSitesProportion(mFac.getInvariantSitesProportion()),
-	_cpijGam(), _rootSequence(mFac.getAlphabet()), _subManager(mFac.getTree()->getNodesNum()),
-	_nodesToSave(nodesToSave), _saveRates(false), _biased_coin(0,1) {
-		// _et = mFac.getTree();
-		// _sp = mFac.getStochasticProcess();
-		// _alph = mFac.getAlphabet();
-
-		size_t alphaSize = _sp->alphabetSize();
-
-		_cpijGam.fillPij(*_et, *_sp);
-		initGillespieSampler();
-		
-
-		std::vector<MDOUBLE> rateProbs;
-		for (int j = 0 ; j < _sp->categories(); ++j) {
-			MDOUBLE currentRateProb = _sp->ratesProb(j);
-			currentRateProb = currentRateProb * (1.0  - _invariantSitesProportion);
-			rateProbs.push_back(currentRateProb);
-		}
-		if (_invariantSitesProportion > 0.0) rateProbs.push_back(_invariantSitesProportion);
-
-		_rateSampler = std::make_unique<DiscreteDistribution>(rateProbs);
-
-		std::vector<MDOUBLE> frequencies;
-		for (int j = 0; j < alphaSize; ++j) {
-			frequencies.push_back(_sp->freq(j));
-		}
-		_frequencySampler = std::make_unique<DiscreteDistribution>(frequencies);
-
-		_simulatedSequences = std::make_unique<sequenceContainer>();
-
-};
-
-void rateMatrixSim::setSaveRates(bool saveRates) {
-	_saveRates = saveRates;
-}
-
-void rateMatrixSim::initGillespieSampler() {
-	_gillespieSampler.resize(_alph->size());
-	for (size_t i = 0; i < _alph->size(); ++i) {
-		std::vector<double> qRates(_alph->size(), 0.0);
-		double sum = -_sp->Qij(i,i);
-		double normalizer = 1.0 / sum;
-		for (size_t j = 0; j < _alph->size(); ++j) {
-			if (i==j) continue;
-			qRates[j] = _sp->Qij(i,j) * normalizer;
-			// std::cout << i << j << "->" << qRates[j] << ",";
-		}
-		// std::cout << "\n" << i << " " << sum << "\n";
-		_gillespieSampler[i] = std::make_unique<DiscreteDistribution>(qRates);
-	}
-}
-
-// simulateTree::simulateTree(const tree&  _inEt,
-// 						   const stochasticProcess& sp,
-// 						   const alphabet* alph) : _sp(sp) {
-// 		_et = _inEt;
-// 		// _sp = sp;
-// 		_alph = alph;
-// 		_avgSubtitutionsPerSite = 0.0;
-// 	};
-
-rateMatrixSim::~rateMatrixSim() {
-}
-
-// void rateMatrixSim::setSeed(size_t seed) {
-// 	_seed = seed;
-// 	_mt_rand->seed(seed);
-// }
-
-void rateMatrixSim::setRng(mt19937_64 *rng) {
-	_mt_rand = rng;
-}
-
-
-// const mt19937_64& rateMatrixSim::getRng(){
-// 	return *_mt_rand;
-// }
-
-
-void rateMatrixSim::generate_substitution_log(int seqLength) {
-	std::vector<MDOUBLE> ratesVec(seqLength);
-
-	MDOUBLE sumOfRatesAcrossSites = 0.0;
-	_rateCategories.resize(seqLength);
-	for (int h = 0; h < seqLength; h++)  {
-		int selectedRandomCategory = _rateSampler->drawSample() - 1;
-		_rateCategories[h] = selectedRandomCategory;
-		if (selectedRandomCategory >= _sp->categories()) {
-			ratesVec[h] = 0.0;
-			continue;
-		}
-		ratesVec[h] = _sp->rates(selectedRandomCategory);
-		sumOfRatesAcrossSites += ratesVec[h];
-	}
-	if (_saveRates) _siteRates.insert(_siteRates.end(), ratesVec.begin(), ratesVec.end());
-	// MDOUBLE sumOfRatesNoramlizingFactor = 1.0 / sumOfRatesAcrossSites;
-
-	// _siteSampler = std::make_unique<DiscreteDistribution>(ratesVec, sumOfRatesNoramlizingFactor);
-	_rootSequence.resize(seqLength);
-	generateRootSeq(seqLength, ratesVec);
-	if ((*_nodesToSave)[_et->getRoot()->id()]) saveSequence(_et->getRoot()->id(), _et->getRoot()->name());
-
-	mutateSeqRecuresively(_et->getRoot(), seqLength);
-	_subManager.clear();
-}
-
-void rateMatrixSim::mutateSeqRecuresively(tree::nodeP currentNode, int seqLength) {
-	if (currentNode->isLeaf()) return;
-
-	for (auto &node: currentNode->getSons()) {
-		mutateSeqAlongBranch(node, seqLength);
-		if ((*_nodesToSave)[node->id()]) saveSequence(node->id(), node->name());
-		mutateSeqRecuresively(node, seqLength);
-		if (!_subManager.isEmpty(node->id())) {
-			_subManager.undoSubs(node->id(), _rootSequence, _rateCategories, _sp.get());
-		}
-	}
-}
-
-void rateMatrixSim::mutateSeqAlongBranch(tree::nodeP currentNode, int seqLength) {
-	const MDOUBLE distToFather = currentNode->dis2father();
-	mutateEntireSeq(currentNode, seqLength);
-	
-	// if (distToFather > 0.5) {
-	// 	mutateEntireSeq(currentNode, seqLength);
-	// } else {
-	// 	mutateSeqGillespie(currentNode, seqLength, distToFather);
-	// }
-	// testSumOfRates();
-}
-
-
-void rateMatrixSim::mutateEntireSeq(tree::nodeP currentNode, int seqLength) {
-	const int nodeId = currentNode->id();
-	const int parentId = currentNode->father()->id();
-
-	for (size_t site = 0; site < seqLength; ++site) {
-		ALPHACHAR parentChar = _rootSequence[site];//_subManager.getCharacter(parentId, site, _rootSequence);
-		if (_rateCategories[site] == _sp->categories()) continue;
-		ALPHACHAR nextChar = _cpijGam.getRandomChar(_rateCategories[site], nodeId, parentChar);
-		if (nextChar != parentChar){
-			_subManager.handleEvent(nodeId, site, nextChar, _rateCategories, _sp.get(), _rootSequence);
-		}
-	}
-}
-
-
-void rateMatrixSim::mutateSeqGillespie(tree::nodeP currentNode, int seqLength, MDOUBLE distToParent) {
-	// std::cout << "mutating sequence using Gillespie!\n";
-
-	const int nodeId = currentNode->id();
-	const int parentId = currentNode->father()->id();
-	MDOUBLE branchLength = distToParent;
-
-	double lambdaParam = _subManager.getReactantsSum();
-	std::exponential_distribution<double> distribution(-lambdaParam);
-	double waitingTime = distribution(*_mt_rand);
-	if (waitingTime < 0) {
-		std::cout << branchLength << " " << lambdaParam << " " << waitingTime << "\n";
-		errorMsg::reportError("waiting time is negative :(");
-	}
-	while (waitingTime < branchLength) {
-		if (waitingTime < 0) {
-			std::cout << branchLength << " " << lambdaParam << " " << waitingTime << "\n";
-			errorMsg::reportError("waiting time is negative :(");
-		}
-
-		int mutatedSite = _subManager.sampleSite(*_mt_rand);
-		ALPHACHAR parentChar = _rootSequence[mutatedSite];
-		ALPHACHAR nextChar = _gillespieSampler[parentChar]->drawSample() - 1;
-		// std::cout << (int)parentChar << "->" << (int)nextChar << "\n";
-		_subManager.handleEvent(nodeId, mutatedSite, nextChar, _rateCategories, _sp.get(), _rootSequence);
-
-		lambdaParam = _subManager.getReactantsSum();
-		branchLength = branchLength - waitingTime;
-		std::exponential_distribution<double> distribution(-lambdaParam);
-		waitingTime = distribution(*_mt_rand);
-
-	}
-}
-
-
-
-
-void rateMatrixSim::generateRootSeq(int seqLength, std::vector<MDOUBLE>& ratesVec) {
-	size_t rootID = _et->getRoot()->id();
-	for (int i = 0; i < seqLength; i++) {
-		ALPHACHAR newChar = _frequencySampler->drawSample() - 1;
-		// ratesVec[i] = ratesVec[i]*(-_sp->Qij(newChar, newChar));
-		_rootSequence[i] =  newChar;
-     }
-	// std::cout << ">Root-sequence\n" << _rootSequence  <<  "\n";
-	// std::cout << ">Rates\n" << ratesVec;
-	_subManager.handleRootSequence(seqLength, ratesVec, _sp.get(), _rootSequence);
-	
-	_rootSequence.setAlphabet(_alph);
-	_rootSequence.setName(_et->getRoot()->name());
-	_rootSequence.setID(_et->getRoot()->id());
-}
-
-
-void rateMatrixSim::saveSequence(const int &nodeId,const std::string &name) {
-	sequence temp(_rootSequence);
-	temp.setName(name);
-	temp.setID(nodeId);
-	// std::cout << temp << "\n";
-	_simulatedSequences->add(temp);
-}
-
-// sequenceContainer rateMatrixSim::toSeqData() {
-// 	sequenceContainer myseqData;
-// 	for (int i=0; i < _simulatedSequences.size(); ++i) {
-// 		myseqData.add(*_simulatedSequences[i]);
-// 	}
-// 	return myseqData;
-// }
-
-
-
-std::unique_ptr<sequenceContainer> rateMatrixSim::getSequenceContainer() {
-	// std::unique_ptr<sequenceContainer> myseqData = std::make_unique<sequenceContainer>();
-	// // sequenceContainer myseqData;
-	// for (int i=0; i < _simulatedSequences.size(); ++i) {
-	// 	tree::nodeP theCurNode = _et->findNodeById(_simulatedSequences[i]->id());
-	// 	if (theCurNode == NULL)
-	// 		errorMsg::reportError("could not find the specified name: " + _simulatedSequences[i]->name());
-	// 	if (theCurNode->isInternal()) continue;
-	auto outputSequences = std::move(_simulatedSequences);
-	_simulatedSequences = std::make_unique<sequenceContainer>();
-	// 	myseqData->add(*std::move(_simulatedSequences[i]));
-	// }
-
-	return outputSequences;
-}
-
-
-bool rateMatrixSim::testSumOfRates() {
-	MDOUBLE sumOfRates = 0.0;
-	for (size_t i = 0; i < _rootSequence.seqLen(); i++) {
-		ALPHACHAR currentChar = _rootSequence[i];
-		MDOUBLE currentQii = _sp->Qij(currentChar, currentChar);
-		MDOUBLE currentRate = _sp->rates(_rateCategories[i]);
-		sumOfRates += (currentQii*currentRate);
-	}
-	MDOUBLE preCalculatedSum = _subManager.getReactantsSum();
-	if (abs(preCalculatedSum - sumOfRates) > 1e-6) {
-		std::cout << "preCalculatedSum=" << preCalculatedSum << " "
-				  << "sumOfRates=" << sumOfRates;
-		errorMsg::reportError("Error in sum of rates calculation!");
-	}
-	std::cout << "preCalculatedSum is correct\n" << "preCalculatedSum=" << preCalculatedSum << " "
-				  << "sumOfRates=" << sumOfRates << "\n";
-
-	return true;
-}