Chris:LabNotes/Moleculo/2013/2013-5-17

From ZhangLabWiki
Jump to navigation Jump to search

Moleculo Data Analysis Part 3[edit]

Results from Test Assembly[edit]

  • Previously I ran test assembly on a small subset of Illumina reads (88179998 reads) that included one 200bp library and one 2kb insert library.
  • There were three tests that I ran:
    • 1a = no error correction for any Illumina read
    • 1b = error correction only on mate-paired 2kb reads
    • 1c = error correction on all Illumina reads
  • Here are the basic assembly statistics:
n       n:200   n:N50   min     N80     N50     N20     max     sum
1772128 504382  181476  200     233     307     456     3238    156.8e6 1a_Illumina_Test_Assembly_Control/test_assembly_control.scafSeq.GC
1768536 535005  189770  200     236     313     471     4194    169.2e6 1b_Illumina_Test_Assembly_no-Mate-correction/test_assembly_1b.scafSeq.GC
1766902 535182  189131  200     236     314     475     4194    169.9e6 1c_Illumina_Test_Assembly_all-correction/test_assembly_1c.scafSeq.GC
  • Analysis: It seems as though having error correction on all reads provided slightly better assembly results, so I'll apply error correction on all Illumina reads for downstream assembly

Test Assembly + Moleculo Reads Assembly[edit]

  • Ran hybrid assembly on genemapster </media/disk-2/cjwei/130419_Oncomelania_Illumina_Moleculo/Triton_Hybrid_Assembly_5_13_2013/test_assembly+moleculo>
  • Commands/instructions to run:
    • Make frg files for combined Illumina test assembly scaffolds (test 1c) + Moleculo data (411,392,248bp)
/media/2TB_store1/cjwei/software/Hybrid_error_correction_assembly_software/wgs_download_2/wgs-7.0/Linux-amd64/bin/fastqToCA -libraryname test_assembly_moleculo -reads test_assembly+moleculo.fastq >test_assembly+moleculo.frq
    • Run assembly
/media/2TB_store1/cjwei/software/Hybrid_error_correction_assembly_software/wgs-7.0/Linux-amd64/bin/runCA -d test_assembly_moleculo -p test_assembly_moleculo test_assembly+moleculo.frg

Test Assembly + Moleculo Reads Assembly Results[edit]

  • Some basic assembly statistics:
n       n:100   n:N50  min    median mean   N50    max    sum 
10322   10322   2812   1000   2458   3094   3682   17558  31.94e6 Illumina_conti
                                                                  g+Moleculo_ass
                                                                  embly/9-
                                                                  terminator/
                                                                  Illumina_conti
                                                                  g+Moleculo.scf
                                                                  .fasta

Test Assembly + Moleculo Reads Assembly Conclusions[edit]

  • From the Illumina-only test assembly, we found that error correction on all read files helped with assembly slightly, so we will implement error correction prior to assembly of all Illumina data
  • From the Illumina+Moleculo assembly, we concluded:
    • Adding Moleculo data drastically increases the scaffold size, while reducing the sum bp assembled (most likely because the shorter Illumina scaffolds are discarded)
    • While the CEGMA results show very few genes predicted from the core eukaryotic gene set, we notice that Illumina+Moleculo data has a few complete genes mapped. This may be because of the longer scaffold size.
  • I'm going to move forward with assembly of all of the Illumina data combined and hybrid assembly of those scaffolds with Moleculo data