-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathslides.html
1023 lines (938 loc) · 37.6 KB
/
slides.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="generator" content="pandoc">
<meta name="author" content="Married with Scaffolds" />
<title>PacBio <3 Illumina</title>
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black-translucent" />
<meta name="viewport" content="width=device-width, initial-scale=1.0, maximum-scale=1.0, user-scalable=no">
<link rel="stylesheet" href="reveal.js/css/reveal.min.css"/>
<style type="text/css">code{white-space: pre;}</style>
<link rel="stylesheet" href="reveal.js/css/theme/simple.css" id="theme">
<link rel="stylesheet" media="print" href="reveal.js/css/print/pdf.css" />
<!--[if lt IE 9]>
<script src="reveal.js/lib/js/html5shiv.js"></script>
<![endif]-->
</head>
<body>
<div class="reveal">
<div class="slides">
<section>
<h1 class="title">PacBio <3 Illumina</h1>
<h2 class="author">Married with Scaffolds</h2>
<h3 class="date">Heinz Ekker (CSF.NGS) 2014-02-11</h3>
</section>
<section id="pacbio-3-illumina" class="slide level1">
<h1>PacBio <3 Illumina</h1>
<h3 id="a-short-very-technical-introduction-to-hybrid-de-novo-genome-assembly-combining-illumina-short-reads-with-pacbio-long-reads.">A short & very technical introduction to hybrid <em>de novo</em> genome assembly combining Illumina short reads with Pacbio long reads.</h3>
<p>Pipeline scripts, markdown source code and data for assembly, analysis and presentation available at <a href="https://github.com/h3kker/assemblyTalk">https://github.com/h3kker/assemblyTalk</a></p>
</section>
<section class="slide level1">
<h2 id="content">Content</h2>
<ol type="1">
<li>An Idiot's Guide to PacBio & Assembly</li>
<li>Hybrid Assembly Strategies</li>
<li>The Data + The Results</li>
<li>Assembly Assessment</li>
<li>Outlook</li>
</ol>
</section>
<section class="slide level1">
<h2 id="pacbio-basics">PacBio Basics</h2>
<figure>
<img src="figure/pacbioSequencing.jpg" alt="Library Preparation + Sequencing" /><figcaption>Library Preparation + Sequencing</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="pacbio-basics-1">PacBio Basics</h2>
<figure>
<img src="figure/subreadFiltering.jpg" alt="Subread Filtering" /><figcaption>Subread Filtering</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="assembly-basics">Assembly Basics</h2>
<figure>
<img src="figure/assemblySteps.jpg" alt="Assembly Steps" /><figcaption>Assembly Steps</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="preprocessing">Preprocessing</h3>
<ul>
<li><p>Illumina Error Correction: K-mer Spectrum (SGA), Suffix Tree/Array, ...</p></li>
<li><p>Error Correction of Pacbio reads with Illumina (details later)</p></li>
<li><p>Adapter trimming</p></li>
<li><p>Quality trimming</p></li>
<li><p>Deduplication</p></li>
</ul>
<p>Some assemblers depend on other, existing tools to perform these steps or do one or more as part of their pipeline. If so, don't use other tools - see Assemblathon.</p>
</section>
<section class="slide level1">
<h3 id="graph-construction">Graph Construction</h3>
<figure>
<img src="figure/drinks.png" alt="assembling hemingway" /><figcaption>assembling hemingway</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-construction-1">Graph Construction</h3>
<figure>
<img src="figure/assemblyGraph.jpg" alt="Suffix-to-Prefix Graph" /><figcaption>Suffix-to-Prefix Graph</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-construction---path-enumeration">Graph Construction - Path Enumeration</h3>
<blockquote>
<p>They stared at the drinks were gone</p>
</blockquote>
<blockquote>
<p>They stared at the drinks went were gone</p>
</blockquote>
<blockquote>
<p>They stared at the drinks the drinks were gone</p>
</blockquote>
<blockquote>
<p>...</p>
</blockquote>
<p>Look for paths that</p>
<ul>
<li>traverse every edge once (Euler)</li>
<li>visit every node once (Hamiltonian)</li>
</ul>
</section>
<section class="slide level1">
<h3 id="graph-construction---path-enumeration-1">Graph Construction - Path Enumeration</h3>
<p>(THEY ST)ARED AT THE DRINKS. THE DRINKS WENT WARM. THEY DRANK.</p>
<figure>
<img src="figure/assemblyPath.jpg" alt="Longest Path" /><figcaption>Longest Path</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-construction---strategies">Graph Construction - Strategies</h3>
<p><strong>Overlap Consensus Layout, eg. SGA</strong></p>
<p>More or less as shown. Minimum length of overlap <em>k</em> is the parameter that determines the graph complexity. Should ideally be as large as the dataset allows (sequencing errors, polyploidity). The ideal assembly visits all nodes exactly once (Hamilton-Path).</p>
<p>String Graphs are a special variant where all transitive edges <code>((X, Y), (X, Z), (Y, Z))</code> are reduced to <code>((X, Z))</code>, <em>irreducible edges</em>.</p>
</section>
<section class="slide level1">
<h3 id="graph-construction---strategies-1">Graph Construction - Strategies</h3>
<p><strong>K-mer based, eg. Abyss, SOAPdenovo</strong></p>
<p>Nodes represent all kmers in the reads. Two kmers are connected if there is a <code>k-1</code> overlap between the nodes (de Bruijn graph). The Euler path that visits each edge exactly once corresponds to a chromosome in an ideal assembly.</p>
<p>K-mer sized (parameter <em>k</em>) should be chosen large enough to reduce the number of wrong connections between contigs, but small enough to allow for errors.</p>
<p><em>Hybrid strategies proposed: Combine contig and graph output from two types of assemblers.</em></p>
</section>
<section class="slide level1">
<h3 id="graph-simplification">Graph Simplification</h3>
<p>Graph structure is very complex due to</p>
<ul>
<li><p>transitive edges like <code>((1,2), (1,3), (2,3))</code></p></li>
<li><p>consecutive nodes like <code>((1,2), (2,3), (3,4))</code></p></li>
<li><p>error reads (branches that converge again later)</p></li>
<li><p>spurious branch points on repeat edges</p></li>
<li><p>dead ends (tips)</p></li>
</ul>
</section>
<section class="slide level1">
<h3 id="graph-simplification---node-merging">Graph Simplification - Node Merging</h3>
<p>Collapse nodes that connect unambiguously (without branching) into one node representing the merged sequence.</p>
<figure>
<img src="figure/assemblyTransitive.jpg" alt="1. Merge Transitive Edges" /><figcaption>1. Merge Transitive Edges</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-simplification---node-merging-1">Graph Simplification - Node Merging</h3>
<p>Collapse nodes that connect unambiguously (without branching) into one node representing the merged sequence.</p>
<figure>
<img src="figure/assemblyConsecutive.jpg" alt="2. Merge Consecutive Nodes" /><figcaption>2. Merge Consecutive Nodes</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-simplification---dead-end-removal">Graph Simplification - Dead End Removal</h3>
<p>Sometimes also: tip erosion. Remove all nodes with connections only in one direction. These can be caused by low coverage regions and read errors. Can also shorten valid contigs!</p>
<figure>
<img src="figure/assemblyDeadEnd.jpg" alt="Dead End" /><figcaption>Dead End</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-simplification---bubble-popping">Graph Simplification - Bubble Popping</h3>
<p>Bubbles due to sequencing errors or polyploid genomes, heterozygosity. Selection of branch based on different criteria like coverage, quality.</p>
<figure>
<img src="figure/assemblyBubble.jpg" alt="Bubble Popping" /><figcaption>Bubble Popping</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="graph-simplification---repeat-tangles">Graph Simplification - Repeat tangles</h3>
<p>Formed in repeated regions, were many reconstructions are possible. Resolved by forming parallel paths. Paired-End constraints can be used to discard invalid edges (too short, too long reconstruction).</p>
<figure>
<img src="figure/assemblyUntangle.jpg" alt="Create parallel paths" /><figcaption>Create parallel paths</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="postprocessing">Postprocessing</h3>
<p><strong>Contigs</strong>: Build contiguous stretches of sequence, filter and correct (consensus)</p>
<p><strong>Scaffolds</strong>: Either with built-in scaffolder or external program. Most assemblers come with their own scaffolder for PE or mate pair library information. Using Pacbio CLRs not yet popular.</p>
<p>Missing sequence information is filled with N (assembly gaps)</p>
</section>
<section class="slide level1">
<h3 id="postprocessing---scaffolding">Postprocessing - Scaffolding</h3>
<p>Use paired end information to join and orient contigs. Can also detect and filter misjoined contigs.</p>
<figure>
<img src="figure/assemblyScaffolding.jpg" alt="Scaffolding" /><figcaption>Scaffolding</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="differences-between-assemblers">Differences between Assemblers</h2>
<p>From the SGA paper:</p>
<blockquote>
<p>[...] We then perform the standard assembly graph post-processing step of removing tips from the graph where a vertex only has a connection in one direction [...]</p>
</blockquote>
<blockquote>
<p>we have developed an algorithm [...] similar to the ‘‘bubble-popping’’ ap- proaches taken by de Bruijn graph assemblers [...]</p>
</blockquote>
<blockquote>
<p>Similar to other approaches to scaffolding (Pop et al. 2004), our method is based on constructing a graph of the relationships between contigs.</p>
</blockquote>
</section>
<section class="slide level1">
<h3 id="differences-between-assemblers-1">Differences between Assemblers</h3>
<p>They all follow the same principles! Main "unique selling points" seem to be algorithms and data structures. The strategies and heuristics employed in graph simplification and postprocessing make the difference in results.</p>
</section>
<section class="slide level1">
<h3 id="differences-between-assemblers-...-and-datasets">Differences between Assemblers ... and datasets</h3>
<figure>
<img src="figure/assemblathon2Cmp.jpg" alt="Bradnam KR et al. Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species. Gigascience. 2013;2(1):10." /><figcaption>Bradnam KR et al. Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species. Gigascience. 2013;2(1):10.</figcaption>
</figure>
</section>
<section id="data-sets-and-assembly-results" class="slide level1">
<h1>Data Sets and Assembly Results</h1>
<p><em>Ustilago bromivora</em>, a fungus with a nice compact genome of about 20Mb.</p>
</section>
<section class="slide level1">
<h3 id="already-sequenced-relatives">Already sequenced relatives:</h3>
<ul>
<li>U. maydis, Corn Smut</li>
</ul>
<p>17.4Mb in 28 contigs.</p>
<p>http://www.broadinstitute.org/annotation/genome/ustilago_maydis.2/Info.html</p>
<ul>
<li>U. hordei, Covered Smut</li>
</ul>
<p>81% of est. 26.1 Mb in 71 supercontigs</p>
<p>25fold coverage with genomic and 10kb paired end library on 454, end-sequencing of a tiled BAC library assigned to 23 chromosomes with optical mapping</p>
<blockquote>
<p>Linning et al., 2004. Genetics 166: 99-111</p>
</blockquote>
<p>http://mips.helmholtz-muenchen.de/genre/proj/MUHDB/About/overview.html</p>
</section>
<section class="slide level1">
<h2 id="libraries">Libraries</h2>
<h3 id="pacbio-clrs">40000 Pacbio CLRs</h3>
<p>mean length of 3910 bp (median 2903, max 20254), see <a href="report.html#toc_2">report.html</a></p>
<h3 id="m-illumina-pe100-reads">42M Illumina PE100 reads</h3>
<p>Library might be problematic. Average insert size estimated at 211bp (+/- 52bp). According to scientist it should be 300-500bp, see <a href="report.html#toc_8">report.html</a></p>
</section>
<section class="slide level1">
<h3 id="illumina-fastqc">Illumina FastQC</h3>
<figure>
<img src="fastqc/440_A_CGTACG_L003.1_fastqc/per_base_quality.png" alt="Quality Scores" /><figcaption>Quality Scores</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="illumina-fastqc-1">Illumina FastQC</h3>
<figure>
<img src="fastqc/440_A_CGTACG_L003.2_fastqc/per_base_sequence_content.png" alt="Read2 Sequence Content" /><figcaption>Read2 Sequence Content</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="illumina-fastqc-2">Illumina FastQC</h3>
<figure>
<img src="fastqc/440_A_CGTACG_L003.1_fastqc/per_sequence_gc_content.png" alt="GC Distribution" /><figcaption>GC Distribution</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="pacbio-fastqc">Pacbio FastQC</h3>
<figure>
<img src="fastqc/filtered_subreads_fastqc/per_base_quality.png" alt="Quality Scores" /><figcaption>Quality Scores</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="pacbio-fastqc-1">Pacbio FastQC</h3>
<figure>
<img src="fastqc/filtered_subreads_fastqc/sequence_length_distribution.png" alt="Length Distribution" /><figcaption>Length Distribution</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="pacbio-fastqc-2">Pacbio FastQC</h3>
<figure>
<img src="fastqc/filtered_subreads_fastqc/per_base_sequence_content.png" alt="Sequence Content" /><figcaption>Sequence Content</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="pacbio-fastqc-3">Pacbio FastQC</h3>
<figure>
<img src="fastqc/filtered_subreads_fastqc/per_sequence_gc_content.png" alt="GC Content" /><figcaption>GC Content</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="sga-preqc">SGA PreQC</h2>
<p>Calculates various metrics to compare with test datasets from Assemblathon2:</p>
<ul>
<li>estimated genome size</li>
<li>branching in de Bruijn graphs with different k-mer sizes</li>
<li>fragment size estimation</li>
<li>kmer spectrum</li>
<li>GC biases</li>
<li>simulated contig length</li>
</ul>
</section>
<section class="slide level1">
<h2 id="preqc">PreQC</h2>
<figure>
<img src="figure/preassembly/variantBranches.png" alt="Variant Branches" /><figcaption>Variant Branches</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="preqc-1">PreQC</h2>
<figure>
<img src="figure/preassembly/repeatBranches.png" alt="Repeat Branches" /><figcaption>Repeat Branches</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="preqc-2">PreQC</h2>
<figure>
<img src="figure/preassembly/errorBranches.png" alt="Error Branches" /><figcaption>Error Branches</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="preqc-3">PreQC</h2>
<figure>
<img src="figure/preassembly/insertSize.png" alt="Insert Size" /><figcaption>Insert Size</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="preqc-4">PreQC</h2>
<figure>
<img src="figure/preassembly/kmerCountDist.png" alt="Kmer Count Distribution" /><figcaption>Kmer Count Distribution</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="preqc-5">PreQC</h2>
<figure>
<img src="figure/preassembly/contigLengths.png" alt="Simulated Contig Lengths" /><figcaption>Simulated Contig Lengths</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="error-correction-approach">Error Correction Approach</h2>
<p>Compensate for high error rate and indel in Pacbio reads by error correcting using relatively accurate Illumina short reads.</p>
<ul>
<li><p>pacBioToCA (based on Celera assembler)</p></li>
<li><p>PreAssembly pipeline (from PacBio SMRTanalysis)</p></li>
</ul>
</section>
<section class="slide level1">
<h3 id="pacbiotoca-error-correction">pacBioToCA Error Correction</h3>
<p>comes with SMRTanalysis software package, but must be run from command line.</p>
<ol type="1">
<li>Create frg file for Illumina reads</li>
<li>Create spec file (by copy&paste...)</li>
<li><p>Run</p>
<p>pacBioToCA -length 500 -partitions 200 -l ec_pacbio -t 16<br />-s pacbio.spec -fastq filtered_subreads.fastq $illuminaFrg</p></li>
<li>wait for ~ 2 days</li>
<li><p>receive a 250MB file called <code>ec_pacbio.fastq</code> and nothing else.</p></li>
</ol>
</section>
<section class="slide level1">
<h3 id="pacbiotoca-error-correction-1">pacBioToCA Error Correction</h3>
<p>Pipeline steps:</p>
<ol type="1">
<li>Create a Gatekeeper store with Illumina reads</li>
<li>OBT: quality trimming, kmer frequency, overlap kmers</li>
<li>OBT: build overlap store</li>
<li>OBT: deduplicate reads (needs lotsa memory)</li>
<li>OBT: trim reads</li>
<li>Overlap: overlap store</li>
<li>Overlap: kmer based error correction</li>
<li>Overlap: Unitigs</li>
<li>ASM: error correction</li>
<li>ASM: Unitigs, create fastq, delete everything else</li>
</ol>
</section>
<section class="slide level1">
<h3 id="pacbiotoca-error-correction-2">pacbioToCA Error Correction</h3>
<p>see <a href="report.html#toc_17">report.html</a></p>
<ul>
<li>Corrected reads are actually shorter than before.</li>
</ul>
<p>(no information about mapping to original reads from pipeline)</p>
<ul>
<li><p>Computationally very intense (good for keeping clusters busy)</p></li>
<li><p>Reduction in Depth makes assembly seem infeasible</p></li>
</ul>
</section>
<section class="slide level1">
<h3 id="smrtanalysis-preassembler-workflow">SMRTanalysis PreAssembler Workflow</h3>
<p>could in theory be run from the web interface, but only with PacBio input (error correcting CLRs with circular consensus reads (CCR). Needs .bas.h5 (primary analysis result from sequencer).</p>
<ol type="1">
<li>start a fake job with only the CLRs from web interface</li>
<li>interrupt, snatch <code>settings.xml</code> and <code>input.xml</code> from job directory</li>
<li><p>run</p>
<p>smrtpipe.py --distribute --output=result/<br /> --params=settings.xml xml:input.xml</p></li>
<li><p>wait 1 - ? days depending on alignment parameters</p></li>
</ol>
</section>
<section class="slide level1">
<h3 id="smrtanalysis-preassembler-workflow-1">SMRTanalysis PreAssembler Workflow</h3>
<p>see <a href="preassembler/workflow/Workflow.summary.html">Pipeline output</a></p>
<ol type="1">
<li>filter subreads, create store for short reads</li>
<li>align short reads to long (14 hours)</li>
<li>layout/consensus (14 hours)</li>
<li>create files</li>
</ol>
<p>Alignment with blasr.</p>
</section>
<section class="slide level1">
<h3 id="smrtanalysis-preassembler-workflow-2">SMRTanalysis PreAssembler Workflow</h3>
<p>see <a href="report.html#toc_32">report.html</a></p>
<ul>
<li><p>Fewer, even shorter reads</p></li>
<li><p>Bad results, but minimal relaxation of alignment criteria produced ~200GB of alignment files which then could not be read</p></li>
<li><p>Very sensitive to parameters for alignment between PacBio and Illumina Reads</p></li>
<li><p>Mapping information between corrected and original reads, better diagnostics</p></li>
</ul>
</section>
<section class="slide level1">
<h3 id="resulting-reads">Resulting Reads</h3>
<figure>
<img src="figure/preAssemblyCumLength.png" alt="cumulative read lengths" /><figcaption>cumulative read lengths</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="hierarchical-assembly">Hierarchical Assembly</h2>
<p>Compensate for short read length by assembling high-fidelity Illumina reads (with high coverage) and resolve repeats and gaps using long Pacbio reads.</p>
<ol type="1">
<li>Run standard assembler</li>
<li>Use Cerulean or PBJelly to scaffold and fill gaps</li>
</ol>
<p>Relatively new, few assemblers have native support for including Pacbio CLRs (in contrast to Mate Pair and Sanger reads)</p>
</section>
<section class="slide level1">
<h3 id="abyss-longscaff">Abyss + Longscaff</h3>
<p>Version 1.3.7 from Dec 11 2013 can use Pacbio CLRs internally with BWA version 0.7.5a+ (with bwa mem support).</p>
<ol type="1">
<li>run <code>abyss-pe</code> with parameter <code>k=64</code> and library paths</li>
<li>watch it crash after initial contigging</li>
<li>run <code>bwa mem</code> manually</li>
<li>restart abyss (hooray makefile!)</li>
<li>receive fasta and graph files (dot) for unitigs, contigs, scaffolds and long-scaff</li>
</ol>
</section>
<section class="slide level1">
<h3 id="abyss-longscaff-1">Abyss + Longscaff</h3>
<blockquote>
<p>Simpson, JT et al. Genome Res 2009</p>
</blockquote>
<ul>
<li>Easy to use (once you get around the bug)</li>
<li>Very fast</li>
</ul>
</section>
<section class="slide level1">
<h3 id="abyss-longscaff-2">Abyss + Longscaff</h3>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">scaffolds</td>
<td style="text-align: right;">698</td>
<td style="text-align: right;">52136</td>
<td style="text-align: right;">200210</td>
</tr>
<tr class="even">
<td style="text-align: left;">longscaff</td>
<td style="text-align: right;">475</td>
<td style="text-align: right;">81601</td>
<td style="text-align: right;">435667</td>
</tr>
</tbody>
</table>
<p>see <a href="report.html#toc_37">report.html</a></p>
</section>
<section class="slide level1">
<h3 id="abyss-cerulean">Abyss + Cerulean</h3>
<ol type="1">
<li>run <code>abyss-pe k=64</code> (without long read library)</li>
<li>align pacbio reads to assembled contigs (<em>not</em> scaffolds?) with blasr</li>
<li>run Cerulean on alignment and <code>${name}-contigs.dot</code></li>
</ol>
</section>
<section class="slide level1">
<h3 id="abyss-cerulean-1">Abyss + Cerulean</h3>
<blockquote>
<p>Deshpande V, et al. Algorithms Bioinformatics 2013</p>
</blockquote>
<ul>
<li>Also quite easy</li>
<li>Not scalable: Crashed on different larger dataset</li>
</ul>
</section>
<section class="slide level1">
<h3 id="abyss-cerulean-2">Abyss + Cerulean</h3>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">scaffolds</td>
<td style="text-align: right;">698</td>
<td style="text-align: right;">52136</td>
<td style="text-align: right;">200210</td>
</tr>
<tr class="even">
<td style="text-align: left;">longscaff</td>
<td style="text-align: right;">475</td>
<td style="text-align: right;">81601</td>
<td style="text-align: right;">435667</td>
</tr>
<tr class="odd">
<td style="text-align: left;">cerulean</td>
<td style="text-align: right;">310</td>
<td style="text-align: right;">106883</td>
<td style="text-align: right;">366413</td>
</tr>
</tbody>
</table>
<p>see <a href="report.html#toc_42">report.html</a></p>
</section>
<section class="slide level1">
<h3 id="sga-assembly">SGA assembly</h3>
<p>String Graph Assembler promises to be more memory-efficient with equally good results. Same first author as abyss.</p>
<ol type="1">
<li>write longish shell script</li>
<li>wait comparatively long</li>
<li>receive error corrected reads fastq, assembly fasta, not much else</li>
</ol>
</section>
<section class="slide level1">
<h3 id="sga-assembly-1">SGA assembly</h3>
<ol type="1">
<li>Create PreQC report</li>
<li>Error Correction using kmer frequencies (3 hours)</li>
<li>Assembly</li>
<li>Scaffolding: Align reads to contigs using BWA</li>
</ol>
</section>
<section class="slide level1">
<h3 id="sga-error-correction">SGA error correction</h3>
<p>discarded ~ 5M reads</p>
<figure>
<img src="figure/sgaCorrectBaseByCycle.png" alt="Base By Cycle" /><figcaption>Base By Cycle</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="sga-assembly-2">SGA assembly</h3>
<blockquote>
<p>Simpson JT. Genome Res. 2012</p>
</blockquote>
<ul>
<li>more complex workflow, more parameters</li>
<li>in-built error correction</li>
</ul>
</section>
<section class="slide level1">
<h3 id="sga-assembly-3">SGA assembly</h3>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">SOAP</td>
<td style="text-align: right;">521</td>
<td style="text-align: right;">78347</td>
<td style="text-align: right;">280862</td>
</tr>
<tr class="even">
<td style="text-align: left;">SGA</td>
<td style="text-align: right;">467</td>
<td style="text-align: right;">57237</td>
<td style="text-align: right;">199401</td>
</tr>
<tr class="odd">
<td style="text-align: left;">Abyss</td>
<td style="text-align: right;">698</td>
<td style="text-align: right;">51236</td>
<td style="text-align: right;">200210</td>
</tr>
</tbody>
</table>
<p>see <a href="report.html#toc_56">report.html</a></p>
</section>
<section class="slide level1">
<h3 id="soap-denovo">SOAP denovo</h3>
<ol type="1">
<li>create protocol file with parameters</li>
<li>write wrapper script for cluster submission</li>
<li>wait (~ 14 hours)</li>
<li>receive graph files (prop) and statistics</li>
</ol>
</section>
<section class="slide level1">
<h3 id="soap-denovo-1">SOAP denovo</h3>
<blockquote>
<p>Luo, R. et al. SOAPdenovo2: an empirically improved memory-efficient short-read de novo assembler. Gigascience 1, 18 (2012).</p>
</blockquote>
<ul>
<li>more complex workflow with more parameters</li>
</ul>
</section>
<section class="slide level1">
<h3 id="soap-denovo-2">SOAP denovo</h3>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">SOAP</td>
<td style="text-align: right;">521</td>
<td style="text-align: right;">78347</td>
<td style="text-align: right;">280862</td>
</tr>
<tr class="even">
<td style="text-align: left;">SGA</td>
<td style="text-align: right;">467</td>
<td style="text-align: right;">57237</td>
<td style="text-align: right;">199401</td>
</tr>
<tr class="odd">
<td style="text-align: left;">Abyss</td>
<td style="text-align: right;">698</td>
<td style="text-align: right;">51236</td>
<td style="text-align: right;">200210</td>
</tr>
</tbody>
</table>
<p>see <a href="report.html#toc_64">report.html</a></p>
</section>
<section class="slide level1">
<h2 id="pbjelly">PBJelly</h2>
<p>Created for filling scaffold gaps.</p>
<ol type="1">
<li>create Protocol.xml with alignment options and cluster parameters</li>
<li>create shell script to run different stages</li>
<li>receive <code>jelly.out.fasta</code> (do NOT run more than one PBJelly per directory!)</li>
</ol>
</section>
<section class="slide level1">
<h2 id="pbjelly-1">PBJelly</h2>
<blockquote>
<p>English AC, et al. PLoS One. 2012</p>
</blockquote>
<ol type="1">
<li>Mapping with <code>blasr</code></li>
<li>find supporting mappings on gap/contig edges</li>
<li>extract sequence information</li>
<li>local assembly of pacbio reads</li>
</ol>
</section>
<section class="slide level1">
<h2 id="pbjelly-2">PBJelly</h2>
<ul>
<li>Some problems: blasr dumped cores for some sequence chunks</li>
<li>Assembly crashed on certain pacbio reads</li>
<li>but results are still good!</li>
</ul>
</section>
<section class="slide level1">
<h2 id="pbjelly-3">PBJelly</h2>
<p>Gap statistics</p>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;">gapped.contigs</th>
<th style="text-align: right;">overall</th>
<th style="text-align: right;">overall.width</th>
<th style="text-align: right;">width.mean</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">cerulean</td>
<td style="text-align: right;">316</td>
<td style="text-align: right;">799</td>
<td style="text-align: right;">529462</td>
<td style="text-align: right;">1675.51</td>
</tr>
<tr class="even">
<td style="text-align: left;">pbj.cerulean</td>
<td style="text-align: right;">152</td>
<td style="text-align: right;">224</td>
<td style="text-align: right;">64066</td>
<td style="text-align: right;">421.49</td>
</tr>
<tr class="odd">
<td style="text-align: left;">sga</td>
<td style="text-align: right;">337</td>
<td style="text-align: right;">612</td>
<td style="text-align: right;">17250</td>
<td style="text-align: right;">51.19</td>
</tr>
<tr class="even">
<td style="text-align: left;">pbj.sga</td>
<td style="text-align: right;">26</td>
<td style="text-align: right;">31</td>
<td style="text-align: right;">927</td>
<td style="text-align: right;">35.65</td>
</tr>
<tr class="odd">
<td style="text-align: left;">soap</td>
<td style="text-align: right;">514</td>
<td style="text-align: right;">3084</td>
<td style="text-align: right;">33891</td>
<td style="text-align: right;">65.94</td>
</tr>
<tr class="even">
<td style="text-align: left;">pbj.soap</td>
<td style="text-align: right;">246</td>
<td style="text-align: right;">2705</td>
<td style="text-align: right;">19088</td>
<td style="text-align: right;">77.59</td>
</tr>
</tbody>
</table>
</section>
<section class="slide level1">
<h2 id="pbjelly-4">PBJelly</h2>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">scaffolds</td>
<td style="text-align: right;">698</td>
<td style="text-align: right;">52136</td>
<td style="text-align: right;">200210</td>
</tr>
<tr class="even">
<td style="text-align: left;">longscaff</td>
<td style="text-align: right;">475</td>
<td style="text-align: right;">81601</td>
<td style="text-align: right;">435667</td>
</tr>
<tr class="odd">
<td style="text-align: left;">cerulean</td>
<td style="text-align: right;">310</td>
<td style="text-align: right;">106883</td>
<td style="text-align: right;">366413</td>
</tr>
<tr class="even">
<td style="text-align: left;">SOAP</td>
<td style="text-align: right;">521</td>
<td style="text-align: right;">78347</td>
<td style="text-align: right;">280862</td>
</tr>
<tr class="odd">
<td style="text-align: left;">SGA</td>
<td style="text-align: right;">467</td>
<td style="text-align: right;">57237</td>
<td style="text-align: right;">199401</td>
</tr>
</tbody>
</table>
<p>Before PBJelly</p>
<table>
<thead>
<tr class="header">
<th style="text-align: left;">set</th>
<th style="text-align: right;"># >2kb</th>
<th style="text-align: right;">N50</th>
<th style="text-align: right;">max</th>
</tr>
</thead>
<tbody>
<tr class="odd">
<td style="text-align: left;">SGA</td>
<td style="text-align: right;">183</td>
<td style="text-align: right;">234931</td>
<td style="text-align: right;">767671</td>
</tr>
<tr class="even">
<td style="text-align: left;">SOAP</td>
<td style="text-align: right;">174</td>
<td style="text-align: right;">201830</td>
<td style="text-align: right;">541843</td>
</tr>
<tr class="odd">
<td style="text-align: left;">Cerulean</td>
<td style="text-align: right;">238</td>
<td style="text-align: right;">159023</td>
<td style="text-align: right;">489237</td>
</tr>
</tbody>
</table>
<p>After PBJelly</p>
</section>
<section class="slide level1">
<h2 id="all-contig-stats">All Contig Stats</h2>
<figure>
<img src="figure/pbjContigs.png" alt="Contig lengths and N50" /><figcaption>Contig lengths and N50</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="all-contig-stats-1">All Contig Stats</h2>
<figure>
<img src="figure/allContigStat.png" alt="Contig lengths and counts" /><figcaption>Contig lengths and counts</figcaption>
</figure>
</section>
<section class="slide level1">
<h2 id="conclusion">Conclusion</h2>
<figure>
<img src="figure/resultsTable.jpg" alt="very subjective results overview" /><figcaption>very subjective results overview</figcaption>
</figure>
</section>
<section id="quality-checks-for-assembly-selection" class="slide level1">
<h1>Quality Checks for Assembly Selection</h1>
</section>
<section class="slide level1">
<h2 id="size-is-not-everything">Size Is Not Everything</h2>
<h2 id="quality-assessment-needed."><strong>Quality Assessment needed.</strong></h2>
<p>But we do not have the luxury of <strong>Assemblathon</strong> or <strong>GAGE</strong> to have a reference to compare to!</p>
<figure>
<img src="figure/sizeVsQuality.png" alt="N50 vs sum of z-scores from different evaluations (Assemblathon 2)" /><figcaption>N50 vs sum of z-scores from different evaluations (Assemblathon 2)</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="alignment-to-close-relative-u.-hordei">Alignment to Close Relative (<em>U. hordei</em>)</h3>
<figure>
<img src="figure/wgsAlignment.png" alt="Nucmer Alignment of SGA Scaffolds to U. hordei assembly" /><figcaption>Nucmer Alignment of SGA Scaffolds to U. hordei assembly</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="alignment-of-pacbio-reads">Alignment of PacBio reads</h3>
<p>Aligned with <code>bwa mem -a -T 60 -k 16 -A 2 -L 4 -t 8 -S -P -k 32</code></p>
<figure>
<img src="figure/bwaMappingRatio.png" alt="matching contigs" /><figcaption>matching contigs</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="alignment-of-pacbio-reads-1">Alignment of PacBio reads</h3>
<figure>
<img src="figure/bwaDepth.png" alt="2D density plot of depth vs contig length" /><figcaption>2D density plot of depth vs contig length</figcaption>
</figure>
<p>A number of contigs with very high depth (>300) were found - A random BLAST produced rDNA.</p>
</section>
<section class="slide level1">
<h3 id="alignment-of-pacbio-reads-2">Alignment of PacBio reads</h3>
<figure>
<img src="figure/bwaAlnDetails.png" alt="Various other results" /><figcaption>Various other results</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="cegma-check-for-presence-of-core-genes">CEGMA: Check for Presence of Core Genes</h3>
<blockquote>
<p>Parra G, et al. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics. 2007;23(9):1061–7</p>
</blockquote>
<ul>
<li>Select 1788 KOGs (eukaryotic orthologous groups) from genes with high identity in organisms from Yeasts to Humans.</li>
<li>Use BLAST to find candidate regions</li>
<li>refine with GeneWise and HMMER</li>
<li>output GFF and report</li>
</ul>
<p>Could be used to examine tentative gene structure!</p>
</section>
<section class="slide level1">
<h3 id="cegma">CEGMA</h3>
<figure>
<img src="figure/cegmaPlot.png" alt="CEGMA results" /><figcaption>CEGMA results</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="reapr">REAPR</h3>
<blockquote>
<p>Hunt M, et al. Genome Biol. 2013</p>
</blockquote>
<figure>
<img src="figure/reaprOutput.png" alt="REAPR output" /><figcaption>REAPR output</figcaption>
</figure>
</section>
<section class="slide level1">
<h3 id="reapr-1">REAPR</h3>
<ul>
<li>align reads back to assembly</li>
<li>infer mismatches and structural errors from paired information (expected insert size distribution)</li>
<li>analyse observed fragment coverage distribution (FCD) vs expected FCD</li>
<li>warn on soft-clipping</li>
</ul>
</section>
<section class="slide level1">
<h4 id="literature">Literature</h4>
<ol type="1">
<li>Simpson JT, Wong K, Jackman SD, et al. ABySS: a parallel assembler for short read sequence data. Genome Res. 2009;19(6):1117–23.</li>
<li>Bradnam KR, Fass JN, Alexandrov A, et al. Assemblathon 2: evaluating de novo methods of genome assembly in three vertebrate species. Gigascience. 2013;2(1):10.</li>
<li>Deshpande V, Fung E, Pham S, Bafna V. Cerulean: A hybrid assembly using high throughput short and long reads. Algorithms Bioinforma. 2013;8126:349–363.</li>
<li>Simpson JT, Durbin R. Efficient de novo assembly of large genomes using compressed data structures. Genome Res. 2012;22(3):549–56.</li>
<li>Simpson J. Exploring Genome Characteristics and Sequence Quality Without a Reference. arXiv Prepr. 2013:1–29.</li>
<li>Salzberg SL, Phillippy AM, Zimin A, et al. GAGE: A critical evaluation of genome assemblies and assembly algorithms. Genome Res. 2012;22(3):557–67.</li>
<li>English AC, Richards S, Han Y, et al. Mind the gap: upgrading genomes with Pacific Biosciences RS long-read sequencing technology. PLoS One. 2012;7(11):e47768.</li>
<li>El-Metwally S, Hamza T, Zakaria M, Helmy M. Next-Generation Sequence Assembly: Four Stages of Data Processing and Computational Challenges Markel S, ed. PLoS Comput. Biol. 2013;9(12):e1003345.</li>
<li>Hunt M, Kikuchi T, Sanders M, et al. REAPR: a universal tool for genome assembly evaluation. Genome Biol. 2013;14(5):R47.</li>
<li>Luo R, Liu B, Xie Y, et al. SOAPdenovo2: an empirically improved memory-efficient short-read de novo assembler. Gigascience. 2012;1(1):18.</li>
<li>Boetzer M, Pirovano W. Toward almost closed genomes with GapFiller. Genome Biol. 2012;13(6):R56.</li>
<li>Parra G, Bradnam K, Korf I. CEGMA: a pipeline to accurately annotate core genes in eukaryotic genomes. Bioinformatics. 2007;23(9):1061–7.</li>
</ol>
</section>
</div>
</div>
<script src="reveal.js/lib/js/head.min.js"></script>
<script src="reveal.js/js/reveal.min.js"></script>