Added new flags for learn. Need to add implementation.
This commit is contained in:
leandrohw
parent
4eb296d0d8
commit
99633f1e9b
1 changed files with 295 additions and 105 deletions
|
|
@ -42,31 +42,34 @@ import edu.utah.ece.async.ibiosim.learn.parameterestimator.ParameterEstimator;
|
|||
* <p>
|
||||
* Options:
|
||||
* <p>
|
||||
* -e: when specified, the program will run parameter estimation.
|
||||
* -l: when specified, parameter estimation will use the estimate the value of the parameters in the list.
|
||||
* -ta [num]: Sets the activation threshold. Default 1.15
|
||||
* -tr [num]: Sets the repression threshold. Default 0.75
|
||||
* -ti [num]: Sets how high a score must be to be considered a parent. Default 0.5
|
||||
* -tm [num]: Sets how close IVs must be in score to be considered for combination. Default 0.01
|
||||
* -tn [num]: Sets minimum number of parents to allow through in SelectInitialParents. Default 2
|
||||
* -tj [num]: Sets the max parents of merged influence vectors, Default 2
|
||||
* -tt [num]: Sets how fast the bound is relaxed for ta and tr, Default 0.025
|
||||
* -d [num]: Sets the debug or output level. Default 0
|
||||
* -wr [num]: Sets how much larger a number must be to be considered as a rise. Default 1
|
||||
* -ws [num]: Sets how far the TSD points are when compared. Default 1
|
||||
* -nb [num]: Sets how many bins are used in the evaluation. Default 4
|
||||
* --lvl: Writes out the suggested levels for every specie
|
||||
* --readLevels: Reads the levels from level.lvl file for every specie
|
||||
* --cpp_harshenBoundsOnTie: Determines if harsher bounds are used when parents tie in CPP.
|
||||
* --cpp_cmp_output_donotInvertSortOrder: Sets the inverted sort order in the 3 places back to normal--cpp_seedParents Determines if parents should be ranked by score, not tsd order in CPP.
|
||||
* --cmp_score_mustNotWinMajority: Determines if score should be used when following conditions are not met a > r+n || r > a + n
|
||||
* --score_donotTossSingleRatioParents: Determines if single ratio parents should be kept
|
||||
* --output_donotTossChangedInfluenceSingleParents: Determines if parents that change influence should not be tossed
|
||||
* -binNumbers: Equal spacing per bin
|
||||
* -noSUCC: to not use successors in calculating probabilities
|
||||
* -PRED: use preicessors in calculating probabilities
|
||||
* -basicFBP: to use the basic FindBaseProb function
|
||||
*
|
||||
* <ul>
|
||||
* <li>-e: when specified, the program will run parameter estimation.</li>
|
||||
* <li>-l: when specified, parameter estimation will use the estimate the value of the parameters in the list.</li>
|
||||
* <li>-ta [num]: Sets the activation threshold. Default 1.15</li>
|
||||
* <li>-tr [num]: Sets the repression threshold. Default 0.75</li>
|
||||
* <li>-ti [num]: Sets how high a score must be to be considered a parent. Default 0.5</li>
|
||||
* <li>-tm [num]: Sets how close IVs must be in score to be considered for combination. Default 0.01</li>
|
||||
* <li>-tn [num]: Sets minimum number of parents to allow through in SelectInitialParents. Default 2</li>
|
||||
* <li>-tj [num]: Sets the max parents of merged influence vectors, Default 2</li>
|
||||
* <li>-tt [num]: Sets how fast the bound is relaxed for ta and tr, Default 0.025</li>
|
||||
* <li>-d [num]: Sets the debug or output level. Default 0</li>
|
||||
* <li>-wr [num]: Sets how much larger a number must be to be considered as a rise. Default 1</li>
|
||||
* <li>-ws [num]: Sets how far the TSD points are when compared. Default 1</li>
|
||||
* <li>-nb [num]: Sets how many bins are used in the evaluation. Default 4</li>
|
||||
* <li>--lvl: Writes out the suggested levels for every species.</li>
|
||||
* <li>--readLevels: Reads the levels from level.lvl file for every species.</li>
|
||||
* <li>--cpp: runs the C++ GeneNet. Default is the Java version. </li>
|
||||
* <li>--cpp_harshenBoundsOnTie: Determines if harsher bounds are used when parents tie in CPP.</li>
|
||||
* <li>--cpp_cmp_output_donotInvertSortOrder: Sets the inverted sort order in the 3 places back to normal</li>
|
||||
* <li>--cpp_seedParents Determines if parents should be ranked by score, not tsd order in CPP.</li>
|
||||
* <li>--cmp_score_mustNotWinMajority: Determines if score should be used when following conditions are not met a > r+n || r > a + n</li>
|
||||
* <li>--score_donotTossSingleRatioParents: Determines if single ratio parents should be kept</li>
|
||||
* <li>--output_donotTossChangedInfluenceSingleParents: Determines if parents that change influence should not be tossed</li>
|
||||
* <li>-binNumbers: Equal spacing per bin</li>
|
||||
* <li>-noSUCC: to not use successors in calculating probabilities</li>
|
||||
* <li>-PRED: use preicessors in calculating probabilities</li>
|
||||
* <li>-basicFBP: to use the basic FindBaseProb function</li>
|
||||
* </ul>
|
||||
*
|
||||
* @author Leandro Watanabe
|
||||
* @author Tramy Nguyen
|
||||
|
|
@ -76,6 +79,31 @@ import edu.utah.ece.async.ibiosim.learn.parameterestimator.ParameterEstimator;
|
|||
*/
|
||||
public class Learn {
|
||||
|
||||
|
||||
private double ta, tr, ti, tm, tn, tj, tt;
|
||||
private int d, wr, ws, nb;
|
||||
private boolean runParameterEstimation, lvl, readLevels, cpp, cpp_harshenBoundsOnTie, cpp_cmp_output_donotInvertSortOrder,
|
||||
cpp_seedParents, cmp_score_mustNotWinMajority, score_donotTossSingleRatioParents, output_donotTossChangedInfluenceSingleParents, binNumbers,
|
||||
noSUCC, PRED, basicFBP;
|
||||
private String directory;
|
||||
private String filename;
|
||||
private List<String> listOfParameters;
|
||||
|
||||
private Learn()
|
||||
{
|
||||
ta = 1.15;
|
||||
tr = 0.75;
|
||||
ti = 0.5;
|
||||
tm = 0.01;
|
||||
tn = 2;
|
||||
tj = 2;
|
||||
tt = 0.025;
|
||||
d = 0;
|
||||
wr = 1;
|
||||
ws = 1;
|
||||
nb = 4;
|
||||
}
|
||||
|
||||
private static void usage() {
|
||||
System.err.println("Description:");
|
||||
System.err.println("\tExecutes bayesian methods for structural learning of regulatory networks using GeneNet or parameter estimation using SRES.");
|
||||
|
|
@ -86,9 +114,8 @@ public class Learn {
|
|||
System.err.println("\t<Project Directory> the directory where the experimental data is located.");
|
||||
System.err.println("Options:");
|
||||
System.err.println("\t-e to execute parameter estimation.");
|
||||
System.err.println("\t-l to specify the list of parameters to estimate. If not specified, all parameters are estimated.");
|
||||
System.err.println("\t to use it, specify the parameters separated by commas (e.g. p1,p2,p3).");
|
||||
|
||||
System.err.println("\t-l to specify the list of parameters to estimate. If not specified, all parameters are estimated. To use it, specify the parameters separated by commas (e.g. p1,p2,p3).");
|
||||
|
||||
System.exit(1);
|
||||
}
|
||||
|
||||
|
|
@ -100,98 +127,261 @@ public class Learn {
|
|||
usage();
|
||||
}
|
||||
|
||||
boolean runParameterEstimation = false;
|
||||
List<String> listOfParameters = null;
|
||||
String filename = args[args.length-2];
|
||||
String directory = args[args.length-1];
|
||||
|
||||
for(int i = 0; i < args.length-2; i++)
|
||||
Learn learn = new Learn();
|
||||
|
||||
learn.filename = args[args.length-2];
|
||||
learn.directory = args[args.length-1];
|
||||
|
||||
int end = args.length-2;
|
||||
for(int i = 0; i < end; i++)
|
||||
{
|
||||
if(args[i].startsWith("-"))
|
||||
String flag = args[i];
|
||||
switch(flag)
|
||||
{
|
||||
if(args[i].length() == 2)
|
||||
case "-e":
|
||||
learn.runParameterEstimation = true;
|
||||
break;
|
||||
case "-l":
|
||||
learn.listOfParameters = new ArrayList<String>();
|
||||
if(i+1 < end)
|
||||
{
|
||||
if(args[i].charAt(1) == 'e')
|
||||
String[] parsedList = args[i+1].split(",");
|
||||
for(String parameter : parsedList)
|
||||
{
|
||||
runParameterEstimation = true;
|
||||
}
|
||||
else if(args[i].charAt(1) == 'l')
|
||||
{
|
||||
if(i+1 < args.length-2)
|
||||
{
|
||||
listOfParameters = new ArrayList<String>();
|
||||
String unparsedList = args[i+1];
|
||||
String[] parsedList = unparsedList.split(",");
|
||||
for(String param : parsedList)
|
||||
{
|
||||
listOfParameters.add(param);
|
||||
}
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
learn.listOfParameters.add(parameter);
|
||||
}
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-ta":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.ta = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-tr":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.tr = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-ti":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.ti = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-tm":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.tm = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-tn":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.tn = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-tj":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.tj = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-tt":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.tt = Double.parseDouble(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-d":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.d = Integer.parseInt(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-wr":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.wr = Integer.parseInt(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-ws":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.ws = Integer.parseInt(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-nb":
|
||||
if(i+1 < end)
|
||||
{
|
||||
learn.nb = Integer.parseInt(args[i+1]);
|
||||
i=i+1;
|
||||
}
|
||||
else
|
||||
{
|
||||
usage();
|
||||
}
|
||||
break;
|
||||
case "-lvl":
|
||||
learn.lvl = true;
|
||||
break;
|
||||
case "--readLevels":
|
||||
learn.readLevels = true;
|
||||
break;
|
||||
case "--cpp":
|
||||
learn.cpp = true;
|
||||
break;
|
||||
case "--cpp_harshenBoundsOnTie":
|
||||
learn.cpp_harshenBoundsOnTie = true;
|
||||
break;
|
||||
case "--cpp_cmp_output_donotInvertSortOrder":
|
||||
learn.cpp_cmp_output_donotInvertSortOrder = true;
|
||||
break;
|
||||
case "--cpp_seedParents":
|
||||
learn.cpp_seedParents = true;
|
||||
break;
|
||||
case "--cmp_score_mustNotWinMajority":
|
||||
learn.cmp_score_mustNotWinMajority = true;
|
||||
break;
|
||||
case "--score_donotTossSingleRatioParents":
|
||||
learn.score_donotTossSingleRatioParents = true;
|
||||
break;
|
||||
case "--output_donotTossChangedInfluenceSingleParents":
|
||||
learn.output_donotTossChangedInfluenceSingleParents = true;
|
||||
break;
|
||||
case "-binNumbers":
|
||||
learn.binNumbers = true;
|
||||
break;
|
||||
case "--noSUCC":
|
||||
learn.noSUCC = true;
|
||||
break;
|
||||
case "--PRED":
|
||||
learn.PRED = true;
|
||||
break;
|
||||
case "--basicFBP":
|
||||
learn.basicFBP = true;
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
if(runParameterEstimation)
|
||||
{
|
||||
try {
|
||||
if(listOfParameters == null)
|
||||
{
|
||||
listOfParameters = new ArrayList<String>();
|
||||
SBMLDocument doc = SBMLReader.read(new File(filename));
|
||||
Model model = doc.getModel();
|
||||
for(Parameter param : model.getListOfParameters())
|
||||
{
|
||||
listOfParameters.add(param.getId());
|
||||
}
|
||||
}
|
||||
|
||||
SpeciesCollection S = new SpeciesCollection();
|
||||
Experiments E = new Experiments();
|
||||
Run.init(filename, S);
|
||||
Run.loadExperiments(directory, S, E);
|
||||
SBMLDocument newDocument = ParameterEstimator.estimate(filename, directory, listOfParameters, E, S);
|
||||
if (newDocument != null)
|
||||
{
|
||||
Model model = newDocument.getModel();
|
||||
for (String parameterId : listOfParameters)
|
||||
{
|
||||
Parameter parameter = model.getParameter(parameterId);
|
||||
if(parameter != null)
|
||||
{
|
||||
System.out.println(parameterId + " = " + parameter.getValue());
|
||||
}
|
||||
else
|
||||
{
|
||||
System.out.println(parameterId + " = NA");
|
||||
}
|
||||
|
||||
}
|
||||
}
|
||||
} catch (IOException e) {
|
||||
System.err.println("The program encoutered IO problems.");
|
||||
} catch (XMLStreamException e) {
|
||||
System.err.println("The program could not parse the input SBML file.");
|
||||
} catch (BioSimException e) {
|
||||
System.err.println("Error: " + e.getMessage());
|
||||
}
|
||||
|
||||
try
|
||||
{
|
||||
if(learn.runParameterEstimation)
|
||||
{
|
||||
learn.runParameterEstimation();
|
||||
}
|
||||
else
|
||||
{
|
||||
learn.runGeneNet();
|
||||
}
|
||||
}
|
||||
catch(Exception e)
|
||||
{
|
||||
System.err.println(e.getMessage());
|
||||
}
|
||||
}
|
||||
|
||||
private void runParameterEstimation() throws XMLStreamException, IOException, BioSimException
|
||||
{
|
||||
if(listOfParameters == null)
|
||||
{
|
||||
listOfParameters = new ArrayList<String>();
|
||||
SBMLDocument doc = SBMLReader.read(new File(filename));
|
||||
Model model = doc.getModel();
|
||||
for(Parameter param : model.getListOfParameters())
|
||||
{
|
||||
listOfParameters.add(param.getId());
|
||||
}
|
||||
}
|
||||
|
||||
SpeciesCollection S = new SpeciesCollection();
|
||||
Experiments E = new Experiments();
|
||||
Run.init(filename, S);
|
||||
Run.loadExperiments(directory, S, E);
|
||||
SBMLDocument newDocument = ParameterEstimator.estimate(filename, directory, listOfParameters, E, S);
|
||||
if (newDocument != null)
|
||||
{
|
||||
Model model = newDocument.getModel();
|
||||
for (String parameterId : listOfParameters)
|
||||
{
|
||||
Parameter parameter = model.getParameter(parameterId);
|
||||
if(parameter != null)
|
||||
{
|
||||
System.out.println(parameterId + " = " + parameter.getValue());
|
||||
}
|
||||
else
|
||||
{
|
||||
System.out.println(parameterId + " = NA");
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private void runGeneNet() throws BioSimException
|
||||
{
|
||||
if(cpp)
|
||||
{
|
||||
|
||||
}
|
||||
else
|
||||
{
|
||||
try {
|
||||
Run.run(filename, directory);
|
||||
} catch (BioSimException e) {
|
||||
System.err.println("Error: " + e.getMessage());
|
||||
}
|
||||
Run.run(filename, directory);
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue