src/hg/makeDb/doc/hg19.txt 1.112

1.112 2010/05/16 20:34:35 hartera
Recreated Signal subtracks for Burge RNA-seq data. Used -bed12 option for bedItemOverlapCount to take the blocks into account.
Index: src/hg/makeDb/doc/hg19.txt
===================================================================
RCS file: /projects/compbio/cvsroot/kent/src/hg/makeDb/doc/hg19.txt,v
retrieving revision 1.111
retrieving revision 1.112
diff -b -B -U 4 -r1.111 -r1.112
--- src/hg/makeDb/doc/hg19.txt	13 May 2010 21:44:19 -0000	1.111
+++ src/hg/makeDb/doc/hg19.txt	16 May 2010 20:34:35 -0000	1.112
@@ -7305,8 +7305,10 @@
 # using reads/per million mapped reads as the data value.
 # 2010-02-17, hartera. Updated trackDb.ra entry to include views.
 # 2010-02-18, hartera. Loaded the bedGraph tables for the Raw Signal
 # subtracks.
+# 2010-05-15 and 2010-05-16, hartera. Re-created the Signal subtracks using 
+# the -bed12 option of bedItemOverlapCount so that blocks are used. 
  
    mkdir /hive/groups/gencode/browser/hg19/burgeRnaSeqGemMapperAlign
    cd /hive/groups/gencode/browser/hg19/burgeRnaSeqGemMapperAlign
 
@@ -7404,31 +7406,36 @@
    end
    # Changed track type in trackDb/human/trackDb.ra to bed 12 and 
    # then did make alpha in trackDb directory.
 
-   # 2010-02-09. Add a Signal track so it is easier to view the data in 
+   # trackDb/human/trackDb.ra entry was updated to include views for Raw Signal
+   # and Alignment subtracks (2010-02-17)
+   # 2010-05-15 and 2010-05-16. Add a Signal track so it is easier to view the data in 
    # regions where there is a high density of reads. 
    cd /hive/data/genomes/hg19/bed/burgeRnaSeqGemMapperAlign
    # Use bedItemOverlapCount to get counts of overlapping items for each base.
    # Need to sort the bed files and then get the number of reads mapped for
    # that tissue. Divide the counts by the number of million mapped reads to 
    # get the number of reads per million mapped reads as the data value. 
+   # Re-make the subtracks using the -bed12 option so that blocks are used 
+   # instead of just the first three fields of the BED file as is the default.
+   rm *.count *.bedGraph
    foreach f (`ls *.bed`)
       echo $f
       set g=$f:r
-      sort $f | bedItemOverlapCount hg19 stdin > ${f}.count
+      sort $f | bedItemOverlapCount -bed12 hg19 stdin > ${f}.count
       set size=`hgsql -Ne "select count(distinct name) from ${g};" hg19`
       awk -v size=${size} 'BEGIN {OFS="\t"} {print $1,$2,$3,($4 / (size/1000000));}' ${f}.count > ${g}.bedGraph
    end
-   # (2010-02-18, hartera)
    # Load the bedGraph tables into the database as Raw Signal tracks.
    foreach f (`ls *.bedGraph`)
       echo $f
       set g=$f:r 
+      hgsql -e "drop table ${g}AllRawSignal;" hg19 
       hgLoadBed -bedGraph=4 hg19 ${g}AllRawSignal $f >>& loadSignal.log
    end
-  # trackDb/human/trackDb.ra entry was updated to include views for Raw Signal
-  # and Alignment subtracks (2010-02-17)
+
+  
 ##########################################################################
 # BUILD ALLEN BRAIN TRACK (DONE 09/30/09 kent)
 
 # Make the working directory