43:
469:
It is quite time-consuming (it can take several days to analyse one chromosome). (Note: In one government lab, running Batman on a set of 100 Agilent Human DNA Methylation Arrays (about 250,000 probes per array) took less than an hour to complete in
Agilent's Genomic Workbench software. Our computer
359:
106:
all the CpGs are methylated). Therefore, to get the full picture of methylation for a given region you have to normalize the signal you get from the MeDIP experiment to the number of CpGs in the region, and this is what the Batman
217:
Based on these assumptions, the signal from the MeDIP channel of the MeDIP-chip or MeDIP-seq experiment depends on the degree of enrichment of DNA fragments overlapping that probe, which in turn depends on the amount of
65:
to isolate methylated DNA sequences. The isolated fragments of DNA are either hybridized to a microarray chip (MeDIP-chip) or sequenced by next-generation sequencing (MeDIP-seq). While this tells you what areas of the
222:, and thus to the number of methylated CpGs on those fragments. In Batman model, the complete dataset from a MeDIP/chip experiment, A, can be represented by a statistical model in the form of the following
127:
The core principle of the Batman algorithm is to model the effects of varying density of CpG dinucleotides, and the effect this has on MeDIP enrichment of DNA fragments. The basic assumptions of Batman:
98:, it will bind both these regions equally and subsequent steps will therefore show equal signals for these two regions. This does not give the full picture of methylation in these two regions (in region
439:) for each tiled region of the genome, then summarizes the most likely methylation state in 100-bp windows by fitting beta distributions to these samples. The modes of the most likely
232:
155:
CpG methylation state is generally highly correlated over hundreds of bases, so CpGs grouped together in 50- or 100-bp windows would have the same methylation state.
382:
419:), that is, the distribution of likely methylation states given one or more sets of MeDIP-chip/MeDIP-seq outputs. To solve this inference problem, Batman uses
592:
Dodge, J.E., Ramsahoye, B.H., Wo, Z.G., Okano, M. & Li, E. De novo methylation of MMLV provirus in embryonic stem cells: CpG versus non-CpG methylation.
636:
190: : total CpG influence parameter, is defined as the sum of coupling factors for any given probe, which provides a measure of local CpG density
24:
487:
One of the basic assumptions of Batman is that all DNA methylation occurs at CpG dinucleotides. While this is generally the case for
119:(i.e. The region is 100% methylated). In this way Batman converts the signals from MeDIP experiments to absolute methylation levels.
139:
Most CpG-poor regions are constitutively methylated while most CpG-rich regions (CpG islands) are constitutively unmethylated.
631:
400:
32:
491:
somatic cells, there are situations where there is widespread non-CpG methylation, such as in plant cells and
223:
354:{\displaystyle f(A\mid m)=\prod _{p}\phi \left(A_{p}\mid A_{\text{base}}+r\sum _{c}C_{cp},\nu ^{-1}\right),}
484:(a loss of 0.4 compared to normal) would have to be multiplied by 1.25 (=2/1.6) to compensate for the loss.
512:
Down, T.A. et al. A Bayesian deconvolution strategy for immunoprecipitation-based DNA methylome analysis.
466:
Because it is non-commercial, there is very little support when using Batman beyond what is in the manual.
460:
142:
There are no fragment biases in MeDIP experiment (approximate range of DNA fragment sizes is 400–700 bp).
473:
219:
404:
70:
are methylated, it does not give absolute methylation levels. Imagine two different genomic regions,
492:
397:
210:
111:
does. Analysing the MeDIP signal of the above example would give Batman scores of 0.5 for region
477:
463:. As such it is not especially user-friendly and is quite a computationally technical process.
440:
367:
451:
It may be useful to take the following points into account when considering using Batman:
420:
28:
95:
57:
MeDIP (methylated DNA immunoprecipitation) is an experimental technique used to assess
625:
42:
87:
83:
36:
94:
has three CpGs, all of which are methylated. As the antibody simply recognizes
537:
488:
424:
202:
146:
213:
since the majority samples used in MeDIP studies contain multiple cell-types.
27:(MeDIP) profiles. It can be applied to large datasets generated using either
533:
108:
476:(CNV) has to be accounted for. For example, the score for a region with a
456:
172:
62:
481:
133:
67:
41:
576:. DNA methylation profiling of human chromosomes 6, 20 and 22.
58:
35:(MeDIP-seq), providing a quantitative estimation of absolute
556:
Bird, A. DNA methylation patterns and epigenetic memory.
167:: coupling factor between probe p and CpG dinucleotide
152:
Only methylated CpGs contribute to the observed signal.
370:
235:
115:(i.e. the region is 50% methylated) and 1 for region
102:
only half the CpGs are methylated, whereas in region
470:
had a 2GHz processor, 24 GB RAM, 64-bit
Windows 7.)
376:
353:
610:Current Topics in Microbiology and Immunology
8:
608:Vanyushin, B.F. DNA methylation in plants.
427:) to generate 100 independent samples from
425:http://www.inference.phy.cam.ac.uk/bayesys/
197: : the methylation status at position
82:has six CpGs (DNA methylation in mammalian
205:in the sample on which it is methylated. m
459:; it is an algorithm performed using the
369:
334:
318:
308:
292:
279:
261:
234:
90:), three of which are methylated. Region
149:are normally distributed with precision.
505:
443:were used as final methylation calls.
23:, is a statistical tool for analysing
17:Bayesian tool for methylation analysis
7:
171:, is defined as the fraction of DNA
536:at base resolution show widespread
201:, which represents the fraction of
637:Applications of Bayesian inference
25:methylated DNA immunoprecipitation
14:
407:techniques can be used to infer
61:methylation levels by using an
39:state in a region of interest.
251:
239:
132:Almost all DNA methylation in
1:
136:happens at CpG dinucleotides.
401:probability density function
159:Basic parameters in Batman:
653:
179:that contain the CpG
33:next-generation sequencing
455:Batman is not a piece of
224:probability distribution
86:generally occurs at CpG
558:Genes & Development
31:arrays (MeDIP-chip) or
378:
355:
49:
632:Computational science
474:Copy number variation
379:
377:{\displaystyle \phi }
356:
175:hybridizing to probe
123:Development of Batman
45:
514:Nature Biotechnology
493:embryonic stem cells
368:
233:
211:continuous variable
209:is considered as a
441:beta distributions
374:
351:
313:
266:
145:The errors on the
50:
583:, 1378–85 (2006).
304:
295:
257:
644:
616:
606:
600:
590:
584:
570:
564:
554:
548:
547:, 315–22 (2009).
526:
520:
519:, 779–85 (2008).
510:
383:
381:
380:
375:
360:
358:
357:
352:
347:
343:
342:
341:
326:
325:
312:
297:
296:
293:
284:
283:
265:
220:antibody binding
19:, also known as
652:
651:
647:
646:
645:
643:
642:
641:
622:
621:
620:
619:
615:, 67–122 (2006)
607:
603:
591:
587:
578:Nature Genetics
571:
567:
555:
551:
527:
523:
511:
507:
502:
449:
421:nested sampling
366:
365:
330:
314:
288:
275:
274:
270:
231:
230:
208:
196:
189:
166:
125:
55:
47:Batman workflow
29:oligonucleotide
12:
11:
5:
650:
648:
640:
639:
634:
624:
623:
618:
617:
601:
585:
565:
563:, 6–21 (2002).
549:
521:
504:
503:
501:
498:
497:
496:
485:
471:
467:
464:
461:command prompt
448:
445:
373:
362:
361:
350:
346:
340:
337:
333:
329:
324:
321:
317:
311:
307:
303:
300:
291:
287:
282:
278:
273:
269:
264:
260:
256:
253:
250:
247:
244:
241:
238:
215:
214:
206:
194:
191:
187:
184:
164:
157:
156:
153:
150:
143:
140:
137:
124:
121:
96:methylated DNA
54:
51:
13:
10:
9:
6:
4:
3:
2:
649:
638:
635:
633:
630:
629:
627:
614:
611:
605:
602:
599:, 41–8 (2002)
598:
595:
589:
586:
582:
579:
575:
572:Eckhardt, F.
569:
566:
562:
559:
553:
550:
546:
543:
540:differences.
539:
535:
531:
525:
522:
518:
515:
509:
506:
499:
494:
490:
486:
483:
479:
475:
472:
468:
465:
462:
458:
454:
453:
452:
446:
444:
442:
438:
434:
430:
426:
422:
418:
414:
410:
406:
402:
399:
395:
391:
387:
371:
348:
344:
338:
335:
331:
327:
322:
319:
315:
309:
305:
301:
298:
289:
285:
280:
276:
271:
267:
262:
258:
254:
248:
245:
242:
236:
229:
228:
227:
225:
221:
212:
204:
200:
192:
185:
182:
178:
174:
170:
162:
161:
160:
154:
151:
148:
144:
141:
138:
135:
131:
130:
129:
122:
120:
118:
114:
110:
105:
101:
97:
93:
89:
88:dinucleotides
85:
84:somatic cells
81:
77:
73:
69:
64:
60:
52:
48:
44:
40:
38:
34:
30:
26:
22:
18:
612:
609:
604:
596:
593:
588:
580:
577:
573:
568:
560:
557:
552:
544:
541:
532:. Human DNA
529:
524:
516:
513:
508:
480:of 1.6 in a
450:
436:
432:
428:
416:
412:
408:
393:
389:
385:
363:
216:
198:
180:
176:
168:
158:
126:
116:
112:
103:
99:
91:
79:
75:
71:
56:
46:
20:
16:
15:
528:Lister, R.
447:Limitations
403:. Standard
203:chromosomes
37:methylation
626:Categories
538:epigenomic
534:methylomes
500:References
489:vertebrate
147:microarray
478:CNV value
372:ϕ
336:−
332:ν
306:∑
286:∣
268:ϕ
259:∏
246:∣
173:molecules
109:algorithm
78:. Region
457:software
405:Bayesian
398:Gaussian
63:antibody
396:) is a
392:,
134:mammals
542:Nature
482:cancer
364:where
68:genome
53:Theory
21:BATMAN
574:et al
530:et al
594:Gene
294:base
74:and
613:301
597:289
545:462
188:tot
59:DNA
628::
581:38
561:16
517:26
226::
165:cp
495:.
437:A
435:|
433:m
431:(
429:f
423:(
417:A
415:|
413:m
411:(
409:f
394:σ
390:μ
388:|
386:x
384:(
349:,
345:)
339:1
328:,
323:p
320:c
316:C
310:c
302:r
299:+
290:A
281:p
277:A
272:(
263:p
255:=
252:)
249:m
243:A
240:(
237:f
207:c
199:c
195:c
193:m
186:C
183:.
181:c
177:p
169:c
163:C
117:B
113:A
104:B
100:A
92:B
80:A
76:B
72:A
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
