Examples with StructureToPolymerChains edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains used on opensource projects

Example 11 with StructureToPolymerChains

use of edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains in project mmtf-spark by sbl-sdsc.

the class PolyPeptideChainStatistics method main.

public static void main(String[] args) throws FileNotFoundException {
    SparkConf conf = new SparkConf().setMaster("local[*]").setAppName(PolyPeptideChainStatistics.class.getSimpleName());
    JavaSparkContext sc = new JavaSparkContext(conf);
    JavaDoubleRDD chainLengths = MmtfReader.readReducedSequenceFile(// read PDB from MMTF-Hadoop sequence file
    sc).flatMapToPair(// split (flatmap) into unique polymer chains
    new StructureToPolymerChains(false, true)).filter(// only consider chains that contain the 20 standard aminoacids
    new PolymerComposition(PolymerComposition.AMINO_ACIDS_20)).mapToDouble(// get the number of groups (residues) in each chain using a lambda expression
    t -> t._2.getNumGroups());
    System.out.println("Protein chains length statistics for proteins in the PDB with the 20 standard amino acids:");
    System.out.println(chainLengths.stats());
    sc.close();
}

Example 12 with StructureToPolymerChains

use of edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains in project mmtf-spark by sbl-sdsc.

the class PdbSequenceToWord2Vec method main.

public static void main(String[] args) throws IOException {
    String path = MmtfReader.getMmtfReducedPath();
    if (args.length != 1) {
        System.err.println("Usage: " + PdbSequenceToWord2Vec.class.getSimpleName() + " <outputFileName>");
        System.exit(1);
    }
    long start = System.nanoTime();
    SparkConf conf = new SparkConf().setMaster("local[*]").setAppName(SecondaryStructureWord2VecEncoder.class.getSimpleName());
    JavaSparkContext sc = new JavaSparkContext(conf);
    // read MMTF Hadoop sequence file and create a non-redundant Pisces
    // subset set (<=40% seq. identity) of L-protein chains
    int sequenceIdentity = 40;
    double resolution = 3.0;
    JavaPairRDD<String, StructureDataInterface> pdb = MmtfReader.readSequenceFile(path, sc).flatMapToPair(new StructureToPolymerChains()).filter(new Pisces(sequenceIdentity, resolution));
    Dataset<Row> data = PolymerSequenceExtractor.getDataset(pdb);
    data.show(10, false);
    // length of polymer sequence segment (number of residues)
    int segmentLength = 11;
    // add Word2Vec encoded feature vector
    ProteinSequenceEncoder encoder = new ProteinSequenceEncoder(data);
    // size of n-grams
    int n = 2;
    int windowSize = (segmentLength - 1) / 2;
    // dimension of vector
    int vectorSize = 50;
    data = encoder.overlappingNgramWord2VecEncode(n, windowSize, vectorSize);
    encoder.getWord2VecModel().save(args[0]);
    long end = System.nanoTime();
    System.out.println(TimeUnit.NANOSECONDS.toSeconds(end - start) + " sec.");
}

Example 13 with StructureToPolymerChains

use of edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains in project mmtf-spark by sbl-sdsc.

the class SecondaryStructureBlosum62Encoder method main.

/**
 * @param args args[0] outputFilePath, args[1] outputFormat (json|parquet)
 * @throws IOException
 * @throws StructureException
 */
public static void main(String[] args) throws IOException {
    String path = MmtfReader.getMmtfReducedPath();
    if (args.length != 2) {
        System.err.println("Usage: " + SecondaryStructureBlosum62Encoder.class.getSimpleName() + " <outputFilePath> + <fileFormat>");
        System.exit(1);
    }
    long start = System.nanoTime();
    SparkConf conf = new SparkConf().setMaster("local[*]").setAppName(SecondaryStructureBlosum62Encoder.class.getSimpleName());
    JavaSparkContext sc = new JavaSparkContext(conf);
    // read MMTF Hadoop sequence file and create a non-redundant Pisces
    // subset set (<=20% seq. identity) of L-protein chains
    int sequenceIdentity = 20;
    double resolution = 3.0;
    JavaPairRDD<String, StructureDataInterface> pdb = MmtfReader.readSequenceFile(path, sc).flatMapToPair(new StructureToPolymerChains()).filter(new Pisces(sequenceIdentity, resolution));
    int segmentLength = 11;
    Dataset<Row> data = SecondaryStructureSegmentExtractor.getDataset(pdb, segmentLength).cache();
    System.out.println("original data     : " + data.count());
    data = data.dropDuplicates("labelQ3", "sequence").cache();
    System.out.println("- duplicate Q3/seq: " + data.count());
    data = data.dropDuplicates("sequence").cache();
    System.out.println("- duplicate seq   : " + data.count());
    // add a property encoded feature vector
    ProteinSequenceEncoder encoder = new ProteinSequenceEncoder(data);
    data = encoder.blosum62Encode();
    data.printSchema();
    data.show(25, false);
    if (args[1].equals("json")) {
        // coalesce data into a single file
        data = data.coalesce(1);
    }
    data.write().mode("overwrite").format(args[1]).save(args[0]);
    long end = System.nanoTime();
    System.out.println(TimeUnit.NANOSECONDS.toSeconds(end - start) + " sec.");
}

Example 14 with StructureToPolymerChains

use of edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains in project mmtf-spark by sbl-sdsc.

the class SecondaryStructureOneHotEncoder method main.

/**
 * @param args args[0] outputFilePath, args[1] outputFormat (json|parquet)
 * @throws IOException
 * @throws StructureException
 */
public static void main(String[] args) throws IOException {
    String path = MmtfReader.getMmtfReducedPath();
    if (args.length < 2) {
        System.err.println("Usage: " + SecondaryStructureOneHotEncoder.class.getSimpleName() + " <outputFilePath> + <fileFormat> + [<modelFileName>]");
        System.exit(1);
    }
    long start = System.nanoTime();
    SparkConf conf = new SparkConf().setMaster("local[*]").setAppName(SecondaryStructureOneHotEncoder.class.getSimpleName());
    JavaSparkContext sc = new JavaSparkContext(conf);
    // read MMTF Hadoop sequence file and create a non-redundant Pisces
    // subset set (<=20% seq. identity) of L-protein chains
    int sequenceIdentity = 20;
    double resolution = 3.0;
    JavaPairRDD<String, StructureDataInterface> pdb = MmtfReader.readSequenceFile(path, sc).flatMapToPair(new StructureToPolymerChains()).filter(new Pisces(sequenceIdentity, resolution));
    // get content
    int segmentLength = 11;
    Dataset<Row> data = SecondaryStructureSegmentExtractor.getDataset(pdb, segmentLength).cache();
    System.out.println("original data     : " + data.count());
    data = data.dropDuplicates("labelQ3", "sequence").cache();
    System.out.println("- duplicate Q3/seq: " + data.count());
    data = data.dropDuplicates("sequence").cache();
    System.out.println("- duplicate seq   : " + data.count());
    // add one-hot encoded sequence feature vector to dataset
    ProteinSequenceEncoder encoder = new ProteinSequenceEncoder(data);
    data = encoder.oneHotEncode();
    data.printSchema();
    data.show(25, false);
    if (args[1].equals("json")) {
        // coalesce data into a single file
        data = data.coalesce(1);
    }
    data.write().mode("overwrite").format(args[1]).save(args[0]);
    long end = System.nanoTime();
    System.out.println(TimeUnit.NANOSECONDS.toSeconds(end - start) + " sec.");
}

Example 15 with StructureToPolymerChains

use of edu.sdsc.mmtf.spark.mappers.StructureToPolymerChains in project mmtf-spark by sbl-sdsc.

the class SecondaryStructurePropertyEncoder method main.

/**
 * @param args outputFilePath outputFormat (json|parquet)
 * @throws IOException
 * @throws StructureException
 */
public static void main(String[] args) throws IOException {
    String path = MmtfReader.getMmtfReducedPath();
    if (args.length != 2) {
        System.err.println("Usage: " + SecondaryStructurePropertyEncoder.class.getSimpleName() + " <outputFilePath> + <fileFormat>");
        System.exit(1);
    }
    long start = System.nanoTime();
    SparkConf conf = new SparkConf().setMaster("local[*]").setAppName(SecondaryStructurePropertyEncoder.class.getSimpleName());
    JavaSparkContext sc = new JavaSparkContext(conf);
    // read MMTF Hadoop sequence file and create a non-redundant Pisces
    // subset set (<=20% seq. identity) of L-protein chains
    int sequenceIdentity = 20;
    double resolution = 3.0;
    JavaPairRDD<String, StructureDataInterface> pdb = MmtfReader.readSequenceFile(path, sc).flatMapToPair(new StructureToPolymerChains()).filter(new Pisces(sequenceIdentity, resolution));
    // get content
    int segmentLength = 11;
    Dataset<Row> data = SecondaryStructureSegmentExtractor.getDataset(pdb, segmentLength).cache();
    System.out.println("original data     : " + data.count());
    data = data.dropDuplicates("labelQ3", "sequence").cache();
    System.out.println("- duplicate Q3/seq: " + data.count());
    data = data.dropDuplicates("sequence").cache();
    System.out.println("- duplicate seq   : " + data.count());
    // add a property encoded feature vector
    ProteinSequenceEncoder encoder = new ProteinSequenceEncoder(data);
    data = encoder.propertyEncode();
    data.printSchema();
    data.show(25, false);
    if (args[1].equals("json")) {
        // coalesce data into a single file
        data = data.coalesce(1);
    }
    data.write().mode("overwrite").format(args[1]).save(args[0]);
    long end = System.nanoTime();
    System.out.println(TimeUnit.NANOSECONDS.toSeconds(end - start) + " sec.");
}