1.2.25

README.md

@@ -1,5 +1,5 @@
-# QoRTs v1.2.19
-(Compiled Fri Feb 17 13:51:51 EST 2017)
+# QoRTs v1.2.25
+(Compiled Sat Mar  4 13:22:59 EST 2017)
 
 The [QoRTs software package](http://hartleys.github.io/QoRTs/) is a fast, efficient, and portable 
 multifunction toolkit designed to assist in

src/HartleyUtils/src/main/scala/internalUtils/commonSeqUtils.scala

@@ -93,10 +93,190 @@ object commonSeqUtils {
     s.map(reverseBaseCharMap(_));
   }
 
+
+
+
+  /******************************************************************************************************
+   * Amino Acid Translation:
+   ******************************************************************************************************/
+
+  val AMINO_ACID_CODON_MAP : scala.collection.immutable.Map[String,String] = scala.collection.immutable.Map[String,String](
+      ("TTT","F"),
+      ("TTC","F"),
+      ("TTA","L"),
+      ("TTG","L"),
+
+      ("CTT","L"),
+      ("CTC","L"),
+      ("CTA","L"),
+      ("CTG","L"),
+
+      ("ATT","I"),
+      ("ATC","I"),
+      ("ATA","I"),
+      ("ATG","M"),
+
+      ("GTT","V"),
+      ("GTC","V"),
+      ("GTA","V"),
+      ("GTG","V"),
+
+      ("TCT","S"),
+      ("TCC","S"),
+      ("TCA","S"),
+      ("TCG","S"),
+
+      ("CCT","P"),
+      ("CCC","P"),
+      ("CCA","P"),
+      ("CCG","P"),
+
+      ("ACT","T"),
+      ("ACC","T"),
+      ("ACA","T"),
+      ("ACG","T"),
+
+      ("GCT","A"),
+      ("GCC","A"),
+      ("GCA","A"),
+      ("GCG","A"),
+
+      ("TAT","Y"),
+      ("TAC","Y"),
+      ("TAA","*"),
+      ("TAG","*"),
+
+      ("CAT","H"),
+      ("CAC","H"),
+      ("CAA","Q"),
+      ("CAG","Q"),
+
+      ("AAT","N"),
+      ("AAC","N"),
+      ("AAA","K"),
+      ("AAG","K"),
+
+      ("GAT","D"),
+      ("GAC","D"),
+      ("GAA","E"),
+      ("GAG","E"),
+
+      ("TGT","C"),
+      ("TGC","C"),
+      ("TGA","*"),
+      ("TGG","W"),
+
+      ("CGT","R"),
+      ("CGC","R"),
+      ("CGA","R"),
+      ("CGG","R"),
+
+      ("AGT","S"),
+      ("AGC","S"),
+      ("AGA","R"),
+      ("AGG","R"),
+
+      ("GGT","G"),
+      ("GGC","G"),
+      ("GGA","G"),
+      ("GGG","G")
+  );
+
+  val AMINO_ACID_ABBRIV_MAP : Map[String,(String,String)] = Map[String,(String,String)](
+      ("F",("Phe","Phenylalanine")),
+      ("L",("Leu","Leucine")),
+      ("I",("Ile","Isoleucine")),
+      ("M",("Met","Methionine")),
+      ("V",("Val","Valine")),
+      ("S",("Ser","Serine")),
+      ("P",("Pro","Proline")),
+      ("T",("Thr","Threonine")),
+      ("A",("Ala","Alanine")),
+      ("Y",("Tyr","Tyrosine")),
+      ("STOP",("STOP","STOP")),
+      ("*",("Stp","STOP")),
+      ("H",("His","Histidine")),
+      ("Q",("Gln","Glutamine")),
+      ("N",("Asn","Asparagine")),
+      ("K",("Lys","Lysine")),
+      ("D",("Asp","AsparticAcid")),
+      ("E",("Glu","GlutamicAcid")),
+      ("C",("Cys","Cysteine")),
+      ("W",("Trp","Tryptophan")),
+      ("R",("Arg","Arginine")),
+      ("G",("Gly","Glycine"))
+      );
+
+  def getAminoAcidFromCode(aa : String, aaNameStyle : Int = 1, strict : Boolean = true) : String = {
+    if(aaNameStyle == 0) return aa;
+    AMINO_ACID_ABBRIV_MAP.get(aa) match {
+      case Some((short,full)) => {
+        if(aaNameStyle == 1) return short;
+        else return full;
+      }
+      case None => {
+        if(strict) error("Error: Unknown amino acid abbriv: "+aa);
+        if(aaNameStyle == 1) return "???";
+        else return "UnknownAminoAcid";
+      }
+    }
+  }
+
+  def getAminoAcidFromCodon(codon : String, aaNameStyle : Int = 1, strict : Boolean = true) : String = {
+    AMINO_ACID_CODON_MAP.get(codon) match {
+      case Some(aa) => {
+        getAminoAcidFromCode(aa,aaNameStyle,strict = strict);
+      }
+      case None => {
+        return getAminoAcidFromCode("?",aaNameStyle,strict = strict);
+      }
+    }
+  }
+
+  def getCodonIterator(iter : Iterator[Char]) : Iterator[String] = {
+    var buffer = "";
+    if(iter.hasNext) buffer = buffer + iter.next;
+    if(iter.hasNext) buffer = buffer + iter.next;
+    if(iter.hasNext) {
+      buffer = buffer + iter.next;
+    } else {
+      return Iterator[String]();
+    }
+
+    if(! iter.hasNext){
+      return Iterator[String](buffer);
+    }
+
+    return new Iterator[String]{
+      var buf = buffer;
+      var hn = true;
+      def hasNext : Boolean = hn;
+      def next : String = {
+        if(iter.hasNext) buffer = "" + iter.next;
+        if(iter.hasNext) buffer = buffer + iter.next;
+        if(iter.hasNext) buffer = buffer + iter.next;
+        if(buffer.size < 3) hn = false;
+        val out = buf;
+        buf = buffer;
+        out;
+      }
+    }
+  }
+
+  def getAminoAcidFromSeq(seq : Iterable[Char], aaNameStyle : Int = 1, strict : Boolean = true) : Iterator[String] = {
+    val iter = getCodonIterator(seq.iterator);
+    iter.map(c => getAminoAcidFromCodon(c,aaNameStyle,strict));
+  }
+
+
+
   /******************************************************************************************************
    * Holder and Helper Classes:
    ******************************************************************************************************/
 
+
+
+
   /******************************************************************************************************
    * FASTQ's:
    */  

src/HartleyUtils/src/main/scala/internalUtils/genomicAnnoUtils.scala

@@ -83,6 +83,7 @@ object genomicAnnoUtils {
     }
   }*/
 
+
   class EfficientGenomeSeqContainer(infiles : Seq[String]){
     var initialReader = internalUtils.fileUtils.getLinesSmartUnzipFromSeq(infiles);
 
@@ -105,7 +106,7 @@ object genomicAnnoUtils {
     //  return (pos - bufferStart) / blockSize;
     //}
 
-    def getSeqForInterval(chrom : String, start : Int, end : Int) : String = {
+    def getSeqForInterval(chrom : String, start : Int, end : Int, truncate : Boolean = false) : String = {
       if(chrom != currChrom) {
         error("ERROR: EfficientGenomeSeqContainer: requested sequence for chromosome other than current chromosome!");
       }
@@ -125,13 +126,6 @@ object genomicAnnoUtils {
       val firstBlockIdx = startPos / blockSize
       val lastBlockIdx = (endPos-1) / blockSize
 
-      /*if(startPos % blockSize < 0 || endPos % blockSize < 0){
-        reportln("Encountered Negative pos?","note");
-        reportBufferStatus;
-        reportln("    start = "+start+", end = "+end+", firstBlockIdx = "+firstBlockIdx+", lastBlockIdx = "+lastBlockIdx,"note");
-        reportln("    startPos % blockSize = "+(startPos % blockSize)+", endPos % blockSize = "+(endPos % blockSize),"note");
-      }*/
-
       val startOffset = startPos % blockSize;
       val endOffset = ((endPos-1) % blockSize) + 1;
 
@@ -143,7 +137,6 @@ object genomicAnnoUtils {
         return buffer(firstBlockIdx).substring(startOffset,blockSize) + buffer.slice(firstBlockIdx+1,lastBlockIdx).mkString("") + buffer(lastBlockIdx).substring(0,endOffset);
       }
       //buffer.substring(start - bufferStart,end - bufferStart);
-
     }
 
     var residualBuffer = "";
@@ -202,10 +195,10 @@ object genomicAnnoUtils {
             error("FATAL ERROR: Cannot find chromosome \""+chrom+"\" in genome FASTA file!")
           }
         }
-
+
+        residualBuffer = "";
         bufferStart = 0;
         buffer = Vector[String]();
-
 
         if(currChrom != chrom){
           reportln("SKIPPING FASTA SEQUENCE FOR CHROMOSOME "+currChrom+" (probable cause: 0 reads found on chromosome, or bam/fasta mis-sorted)!","note");
@@ -240,11 +233,30 @@ object genomicAnnoUtils {
       maxBuffer = math.max(maxBuffer,buffer.length);
     }
 
+    def extendBufferToOrTruncate(end : Int) : Int = {
+      while(currIter.hasNext && bufferEnd < end){
+        buffer = buffer :+ getNextBlock();
+      }
+      bufferEnd;
+    }
+
     def reportBufferStatus {
       reportln("   [GenomeSeqContainer Status: "+"buf:("+currChrom+":"+bufferStart+"-"+bufferEnd+") "+"n="+buffer.length+", "+"MaxSoFar="+maxBuffer+"]","debug");
     }
+
+    def fullStatusDump() : String = {
+      return "EfficientGenomeSeqContainer.fullStatusDump(): \n"+
+             "currChrom = " +currChrom + "\n"+
+             "bufferStart = " +bufferStart + "\n"+
+             "bufferEnd = "+bufferEnd+"\n"+
+             "buffer.size = "+buffer.size+"\n"+
+             "residualBuffer = " +residualBuffer + "\n"+
+             "buffer:\n"+
+             buffer.mkString("\n");
+    }
   }
 
+
   class EfficientGenomeSeqContainer_OLD(infiles : Seq[String]){
     val initialReader = internalUtils.fileUtils.getLinesSmartUnzipFromSeq(infiles);
 

src/HartleyUtils/src/main/scala/runner/runner.scala

@@ -9,9 +9,9 @@ import internalUtils.commandLineUI._;
 
 object runner {
 
-  final val QORTS_VERSION = "1.2.19"; // REPLACE_THIS_QORTS_VERSION_VARIABLE_WITH_VERSION_NUMBER          (note this exact text is used in a search-and-replace. Do not change it.)
-  final val QORTS_COMPILE_DATE = "Fri Feb 17 13:51:51 EST 2017"; // REPLACE_THIS_QORTS_DATE_VARIABLE_WITH_DATE          (note this exact text is used in a search-and-replace. Do not change it.)
-  final val QORTS_COMPILE_TIME : Long = 1487357511; // REPLACE_THIS_QORTS_DATE_VARIABLE_WITH_TIME          (note this exact text is used in a search-and-replace. Do not change it.)
+  final val QORTS_VERSION = "1.2.25"; // REPLACE_THIS_QORTS_VERSION_VARIABLE_WITH_VERSION_NUMBER          (note this exact text is used in a search-and-replace. Do not change it.)
+  final val QORTS_COMPILE_DATE = "Sat Mar  4 13:22:59 EST 2017"; // REPLACE_THIS_QORTS_DATE_VARIABLE_WITH_DATE          (note this exact text is used in a search-and-replace. Do not change it.)
+  final val QORTS_COMPILE_TIME : Long = 1488651779; // REPLACE_THIS_QORTS_DATE_VARIABLE_WITH_TIME          (note this exact text is used in a search-and-replace. Do not change it.)
 
   final val QORTS_MAJOR_VERSION = QORTS_VERSION.split("\\.")(0);
   final val QORTS_MINOR_VERSION = QORTS_VERSION.split("\\.")(1);

src/QoRTs/DESCRIPTION

@@ -1,6 +1,6 @@
 Package: QoRTs
-Version: 1.2.19
-Date: 2017-02-17
+Version: 1.2.25
+Date: 2017-03-04
 Title: Quality of RNA-seq Tool
 Authors@R: c(person("Stephen Hartley, PhD", "Developer", role = c("aut", "cre"),
                      email = "[email protected]"))