-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathDA.html
589 lines (510 loc) · 18.6 KB
/
DA.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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8" />
<meta name="generator" content="pandoc" />
<meta http-equiv="X-UA-Compatible" content="IE=EDGE" />
<title>Differential Abundance Testing</title>
<script src="site_libs/header-attrs-2.27/header-attrs.js"></script>
<script src="site_libs/jquery-3.6.0/jquery-3.6.0.min.js"></script>
<meta name="viewport" content="width=device-width, initial-scale=1" />
<link href="site_libs/bootstrap-3.3.5/css/cosmo.min.css" rel="stylesheet" />
<script src="site_libs/bootstrap-3.3.5/js/bootstrap.min.js"></script>
<script src="site_libs/bootstrap-3.3.5/shim/html5shiv.min.js"></script>
<script src="site_libs/bootstrap-3.3.5/shim/respond.min.js"></script>
<style>h1 {font-size: 34px;}
h1.title {font-size: 38px;}
h2 {font-size: 30px;}
h3 {font-size: 24px;}
h4 {font-size: 18px;}
h5 {font-size: 16px;}
h6 {font-size: 12px;}
code {color: inherit; background-color: rgba(0, 0, 0, 0.04);}
pre:not([class]) { background-color: white }</style>
<script src="site_libs/navigation-1.1/tabsets.js"></script>
<link href="site_libs/highlightjs-9.12.0/default.css" rel="stylesheet" />
<script src="site_libs/highlightjs-9.12.0/highlight.js"></script>
<style type="text/css">
code{white-space: pre-wrap;}
span.smallcaps{font-variant: small-caps;}
span.underline{text-decoration: underline;}
div.column{display: inline-block; vertical-align: top; width: 50%;}
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
ul.task-list{list-style: none;}
</style>
<style type="text/css">code{white-space: pre;}</style>
<script type="text/javascript">
if (window.hljs) {
hljs.configure({languages: []});
hljs.initHighlightingOnLoad();
if (document.readyState && document.readyState === "complete") {
window.setTimeout(function() { hljs.initHighlighting(); }, 0);
}
}
</script>
<style type = "text/css">
.main-container {
max-width: 940px;
margin-left: auto;
margin-right: auto;
}
img {
max-width:100%;
}
.tabbed-pane {
padding-top: 12px;
}
.html-widget {
margin-bottom: 20px;
}
button.code-folding-btn:focus {
outline: none;
}
summary {
display: list-item;
}
details > summary > p:only-child {
display: inline;
}
pre code {
padding: 0;
}
</style>
<style type="text/css">
.dropdown-submenu {
position: relative;
}
.dropdown-submenu>.dropdown-menu {
top: 0;
left: 100%;
margin-top: -6px;
margin-left: -1px;
border-radius: 0 6px 6px 6px;
}
.dropdown-submenu:hover>.dropdown-menu {
display: block;
}
.dropdown-submenu>a:after {
display: block;
content: " ";
float: right;
width: 0;
height: 0;
border-color: transparent;
border-style: solid;
border-width: 5px 0 5px 5px;
border-left-color: #cccccc;
margin-top: 5px;
margin-right: -10px;
}
.dropdown-submenu:hover>a:after {
border-left-color: #adb5bd;
}
.dropdown-submenu.pull-left {
float: none;
}
.dropdown-submenu.pull-left>.dropdown-menu {
left: -100%;
margin-left: 10px;
border-radius: 6px 0 6px 6px;
}
</style>
<script type="text/javascript">
// manage active state of menu based on current page
$(document).ready(function () {
// active menu anchor
href = window.location.pathname
href = href.substr(href.lastIndexOf('/') + 1)
if (href === "")
href = "index.html";
var menuAnchor = $('a[href="' + href + '"]');
// mark the anchor link active (and if it's in a dropdown, also mark that active)
var dropdown = menuAnchor.closest('li.dropdown');
if (window.bootstrap) { // Bootstrap 4+
menuAnchor.addClass('active');
dropdown.find('> .dropdown-toggle').addClass('active');
} else { // Bootstrap 3
menuAnchor.parent().addClass('active');
dropdown.addClass('active');
}
// Navbar adjustments
var navHeight = $(".navbar").first().height() + 15;
var style = document.createElement('style');
var pt = "padding-top: " + navHeight + "px; ";
var mt = "margin-top: -" + navHeight + "px; ";
var css = "";
// offset scroll position for anchor links (for fixed navbar)
for (var i = 1; i <= 6; i++) {
css += ".section h" + i + "{ " + pt + mt + "}\n";
}
style.innerHTML = "body {" + pt + "padding-bottom: 40px; }\n" + css;
document.head.appendChild(style);
});
</script>
<!-- tabsets -->
<style type="text/css">
.tabset-dropdown > .nav-tabs {
display: inline-table;
max-height: 500px;
min-height: 44px;
overflow-y: auto;
border: 1px solid #ddd;
border-radius: 4px;
}
.tabset-dropdown > .nav-tabs > li.active:before, .tabset-dropdown > .nav-tabs.nav-tabs-open:before {
content: "\e259";
font-family: 'Glyphicons Halflings';
display: inline-block;
padding: 10px;
border-right: 1px solid #ddd;
}
.tabset-dropdown > .nav-tabs.nav-tabs-open > li.active:before {
content: "\e258";
font-family: 'Glyphicons Halflings';
border: none;
}
.tabset-dropdown > .nav-tabs > li.active {
display: block;
}
.tabset-dropdown > .nav-tabs > li > a,
.tabset-dropdown > .nav-tabs > li > a:focus,
.tabset-dropdown > .nav-tabs > li > a:hover {
border: none;
display: inline-block;
border-radius: 4px;
background-color: transparent;
}
.tabset-dropdown > .nav-tabs.nav-tabs-open > li {
display: block;
float: none;
}
.tabset-dropdown > .nav-tabs > li {
display: none;
}
</style>
<!-- code folding -->
</head>
<body>
<div class="container-fluid main-container">
<div class="navbar navbar-default navbar-fixed-top" role="navigation">
<div class="container">
<div class="navbar-header">
<button type="button" class="navbar-toggle collapsed" data-toggle="collapse" data-bs-toggle="collapse" data-target="#navbar" data-bs-target="#navbar">
<span class="icon-bar"></span>
<span class="icon-bar"></span>
<span class="icon-bar"></span>
</button>
<a class="navbar-brand" href="index.html">Microbiome Data Analysis</a>
</div>
<div id="navbar" class="navbar-collapse collapse">
<ul class="nav navbar-nav">
<li>
<a href="about.html">Home</a>
</li>
<li>
<a href="pkgs.html">Installation</a>
</li>
<li class="dropdown">
<a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" data-bs-toggle="dropdown" aria-expanded="false">
Descriptives
<span class="caret"></span>
</a>
<ul class="dropdown-menu" role="menu">
<li>
<a href="prepro.html">Preprocessing</a>
</li>
<li>
<a href="alphadiv.html">alpha diversity</a>
</li>
<li>
<a href="betadiv.html">beta diversity</a>
</li>
</ul>
</li>
<li class="dropdown">
<a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" data-bs-toggle="dropdown" aria-expanded="false">
Statistical inference
<span class="caret"></span>
</a>
<ul class="dropdown-menu" role="menu">
<li>
<a href="DA.html">Differential Abundance</a>
</li>
<li>
<a href="DAmixed.html">Differential Abundance - Repeated measures</a>
</li>
<li>
<a href="adonistest.html">ANOVA on beta diversity</a>
</li>
</ul>
</li>
<li class="dropdown">
<a href="#" class="dropdown-toggle" data-toggle="dropdown" role="button" data-bs-toggle="dropdown" aria-expanded="false">
Supervised Multivariate models
<span class="caret"></span>
</a>
<ul class="dropdown-menu" role="menu">
<li>
<a href="hm.html">Heatmaps</a>
</li>
<li>
<a href="cca.html">Canonical Correlation Analysis</a>
</li>
<li>
<a href="mediationanalysis.html">Mediation analysis via Propensity Scoring</a>
</li>
</ul>
</li>
<li>
<a href="https://github.com/mortenarendt/MicrobiomeDataAnalysis/tree/master/data">Data</a>
</li>
</ul>
<ul class="nav navbar-nav navbar-right">
</ul>
</div><!--/.nav-collapse -->
</div><!--/.container -->
</div><!--/.navbar -->
<div id="header">
<h1 class="title toc-ignore">Differential Abundance Testing</h1>
</div>
<p>Having a study design in with there are two (or more) groups of
samples, say male/female, diet1/diet2, etc. one is often interested in
knowing:</p>
<ul>
<li>Is there a difference between the groups?</li>
<li>Which bacteria / OTUs are most different between the groups?</li>
</ul>
<p>This concept is within microbiome statistics refered to as
<strong>Differential Abundance Testing</strong>.</p>
<p>The principal behind is very simple: Given an OTU matrix of
<em>p</em> different OTUs;</p>
<ol style="list-style-type: decimal">
<li><p>Perform <em>p</em> univariate tests assigning an effect size and
a p-value to each OTU on the question of differential abundance. That
could for example be based on a simple t-test</p></li>
<li><p>Arrange the <em>p</em> tests according to the p-value going from
the smallest (most different OTU) to the largest (least different
OTU)</p></li>
<li><p>Deside a cut point for which OTUs to be assigned
<em>discoveries</em>. I.e. OTUs with p-values below <span
class="math inline">\(p_{cut}\)</span> to be discoveries. This task is
known as multiple-testing-correction, and is a general statistical
concept.</p></li>
</ol>
<p>Given that todays microbiome studies generates a large number of
OTUs, <em>p</em> often higher than 1.000 there are two concerns:</p>
<ul>
<li>Choose a reliable and powerful univariate model</li>
<li>Boost the discovery chance by filter and/or aglommeration</li>
</ul>
<p>The paper by Thorsen et al 2016 investigates a range of statistical
strategies on both the ability to sort the OTUs from the most
discriminatory to the least, as well as the false discovery rate
control. Russel et al (2019) have streamlined this approach for
selecting the relevant statistical engine, which nicely screens +20
methods in very few lines of code (see <a
href="https://github.com/Russel88/DAtest">DAtest</a> for code)</p>
<p>(see refs below)</p>
<div id="exercise" class="section level1" number="1">
<h1><span class="header-section-number">1</span> Exercise</h1>
<div id="setting-up-data" class="section level2" number="1.1">
<h2><span class="header-section-number">1.1</span> Setting up data</h2>
<p>Rarefy the <em>mice birth mode</em> data to even depth, and use the
phyloseq_to_deseq2() function to convert format including birht mode as
class information.</p>
</div>
<div id="deseq" class="section level2" number="1.2">
<h2><span class="header-section-number">1.2</span> DEseq</h2>
<p>Perform DA test using DESeq() from the DESeq2 library. You might need
to install this library in advance.</p>
<pre class="r"><code>library(phyloseq)
library(DESeq2)
load('./data/Mice_csec.RData')
birhtmode_ds <- phyloseq_to_deseq2(phyX, ~ Birth_mode)
res <- DESeq(birhtmode_ds, test="Wald", fitType="parametric")</code></pre>
</div>
<div id="interpret-results" class="section level2" number="1.3">
<h2><span class="header-section-number">1.3</span> Interpret
results</h2>
<p>Extract results and combine with tax information (from tax_table())
as well as read frequency per OTU as well as presence/absence
percentage.</p>
<pre class="r"><code>tb <- results(res, cooksCutoff = FALSE)
txtb <- data.frame(tax_table(phyX))
abu <- taxa_sums(phyX)
presence <- apply(otu_table(phyX)>0,1,sum)
df_res <- data.frame(est = tb$log2FoldChange, se = tb$lfcSE, pv = tb$pvalue, pvadj = tb$padj, abu, presence)
df_res <- cbind(df_res, txtb)
df_res$name <- rownames(df_res)</code></pre>
</div>
<div id="is-this-different-from-null" class="section level2"
number="1.4">
<h2><span class="header-section-number">1.4</span> Is this different
from <em>null</em>?</h2>
<p>Plot a histogram of the unadjusted p-values. How should this plot
look like if there were <em>no</em> birth mode information in the
data?</p>
</div>
<div id="vulcano-plot" class="section level2" number="1.5">
<h2><span class="header-section-number">1.5</span> Vulcano plot</h2>
<p>Plot a vulcano plot of the results. Try to facet out / or color
according to a taxonomic level to get deeper insight on which bacteria
that are affected the most.</p>
<pre class="r"><code>ggplot(data = df_res, aes(est,-log10(pv), color = Rank4)) +
geom_point() +
geom_text(data = df_res[df_res$pv<0.0001,], aes(label = name), color = 'black') + # put label on the top stuff
theme(legend.position = 'none') +
facet_wrap(~Rank3)</code></pre>
</div>
<div id="rarecommon-vs-difference" class="section level2" number="1.6">
<h2><span class="header-section-number">1.6</span> rare/common vs
difference</h2>
<p>Interpret the results as a function of read-frequency or
presence/absence. Is it the dominating species that are most
different?</p>
</div>
<div id="effect-of-agglomoration" class="section level2" number="1.7">
<h2><span class="header-section-number">1.7</span> Effect of
agglomoration</h2>
<p>Perform tax glommeration (tax_glom()) at some taxonomic level, and
re-run the analysis. Do you get the same results?</p>
</div>
<div id="metagenomeseq" class="section level2" number="1.8">
<h2><span class="header-section-number">1.8</span> Metagenomeseq</h2>
<p>Do exactly the same just using t-test, wilcox tests as well as
metagenomeseq’s featureModel() and compare the results? Below are some
code-inspiration.</p>
<pre class="r"><code>library(tidyverse)
library(broom)
library(metagenomeSeq)
mgsdata <- phyloseq_to_metagenomeSeq(phyX)
mgsdata <- cumNorm(mgsdata, p=.5)
mod <- model.matrix(~1+Birth_mode, data=pData(mgsdata))
mgsfit <- fitFeatureModel(obj = mgsdata,mod = mod)
predictor <- sample_data(phyX)$Birth_mode
# do some preprocessing
count_table <- log(t(otu_table(phyX)) + 1)
count_table <- apply(count_table, 2, function(x) x/sum(x))
# use tidyverse and broom to get test
tb_ttest <- count_table %>%
cbind(predictor) %>%
data.frame() %>%
gather(otu,ra,-predictor) %>%
group_by(otu) %>%
do(t.test(data = ., ra~predictor) %>% tidy)</code></pre>
<pre class="r"><code>allres <- bind_rows(
data.frame(estimate = mgsfit@fitZeroLogNormal$logFC,
p.value = mgsfit@pvalues,
otu = mgsfit@taxa,
test = 'metagenomeseq'),
data.frame(estimate = tb$log2FoldChange,
p.value = tb$pvalue,
otu = rownames(tb),
test = 'DESeq2'),
tb_ttest %>% mutate(test = 't-test')
)
#plot it!
ggplot(data = allres, aes(estimate,-log10(p.value), label = otu)) +
geom_point() +
geom_text(data = allres[allres$p.value<10e-7 & !is.na(allres$p.value),]) +
facet_wrap(~test,nrow = 1, scales = 'free_x')</code></pre>
<p><img src="DA_files/figure-html/unnamed-chunk-5-1.png" width="672" /></p>
</div>
<div id="put-infeerence-on-the-phylogentic-tree" class="section level2"
number="1.9">
<h2><span class="header-section-number">1.9</span> Put infeerence on the
phylogentic tree</h2>
<p>Because the number of taxa is larger, we restrict the plot to include
the once with a somewhat significant signal.</p>
<pre class="r"><code>library(ggtree)
# Filter object and inference data.
dfsel <- df %>%
dplyr::filter(pv_MGS<0.1) %>%
dplyr::mutate(logFC = ifelse(abs(est_MGS)>1.5, sign(est_MGS)*1.5,est_MGS))
phyobj <- phyX %>%
subset_taxa(rownames(phyX@tax_table) %in% rownames(dfsel)) # this way to make sure it is not depend on order.
TREE <- phy_tree(phyobj)
TXtab <- as.data.frame(tax_table(phyobj))
# merge on inferential stats
AA <- TXtab %>% rownames_to_column('otu') %>%
left_join(dfsel%>% rownames_to_column('otu') , by = 'otu')
# initiale tree
g3 <- ggtree(TREE,layout = 'circular',branch.length="none")
g3 <- g3 %<+% AA
g3 <- g3 +
geom_tippoint(aes(x = x+1, fill = logFC, subset = isTip), shape=21,size = 2 ) +
# geom_text(aes(label = node)) + # use this to get node numbers, and then apply in geom_highlight
geom_hilight(node=1038, fill="gray10", alpha=.2) +
scale_fill_gradient2(low = 'red',high = 'darkgreen',
midpoint = 0,mid = 'white',
na.value = 'grey95',
name = 'logFC') +
theme(legend.position="right",legend.title=element_blank())
g3</code></pre>
<p><img src="DA_files/figure-html/unnamed-chunk-9-1.png" width="960" /></p>
<p>Maybe save it instead of plotting it in R.:</p>
<pre class="r"><code>ggsave('cladogramwithinference.pdf',g3,height = 20, width = 20)</code></pre>
</div>
</div>
<div id="references" class="section level1 unnumbered">
<h1 class="unnumbered">References</h1>
<p><a
href="https://microbiomejournal.biomedcentral.com/articles/10.1186/s40168-016-0208-8">Thorsen,
Jonathan, Asker Brejnrod, Martin Mortensen, Morten A. Rasmussen, Jakob
Stokholm, Waleed Abu Al-Soud, Søren Sørensen, Hans Bisgaard, and
Johannes Waage. <strong>Large-scale benchmarking reveals false
discoveries and count transformation sensitivity in 16S rRNA gene
amplicon data analysis methods used in microbiome studies</strong>,
<em>Microbiome</em> 4, no. 1 (2016): 62.</a></p>
<p><a
href="http://www.nature.com/nmeth/journal/v10/n12/full/nmeth.2658.html">Paulson,
Joseph N., O. Colin Stine, Héctor Corrada Bravo, and Mihai Pop.
<strong>Differential abundance analysis for microbial marker-gene
surveys</strong>. Nature methods 10, no. 12 (2013): 1200.</a></p>
<p><a
href="https://genomebiology.biomedcentral.com/articles/10.1186/s13059-014-0550-8">Love,
Michael I., Wolfgang Huber, and Simon Anders. <strong>Moderated
estimation of fold change and dispersion for RNA-seq data with
DESeq2</strong>. Genome biology 15.12 (2014): 550.</a></p>
<p><a
href="https://www.biorxiv.org/content/biorxiv/early/2018/01/02/241802.full.pdf">Russel,
Jakob, Jonathan Thorsen, Asker D. Brejnrod, Hans Bisgaard, Soren J.
Sorensen, and Mette Burmolle. <strong>DAtest: a framework for choosing
differential abundance or expression method</strong>. bioRxiv (2018):
241802.</a></p>
<p>Xia, Yinglin, Jun Sun, and Ding-Geng Chen. <strong>Statistical
analysis of microbiome data with R</strong>. Springer, 2018.</p>
</div>
</div>
<script>
// add bootstrap table styles to pandoc tables
function bootstrapStylePandocTables() {
$('tr.odd').parent('tbody').parent('table').addClass('table table-condensed');
}
$(document).ready(function () {
bootstrapStylePandocTables();
});
</script>
<!-- tabsets -->
<script>
$(document).ready(function () {
window.buildTabsets("TOC");
});
$(document).ready(function () {
$('.tabset-dropdown > .nav-tabs > li').click(function () {
$(this).parent().toggleClass('nav-tabs-open');
});
});
</script>
<!-- code folding -->
<!-- dynamically load mathjax for compatibility with self-contained -->
<script>
(function () {
var script = document.createElement("script");
script.type = "text/javascript";
script.src = "https://mathjax.rstudio.com/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML";
document.getElementsByTagName("head")[0].appendChild(script);
})();
</script>
</body>
</html>