P281424/AMR
1
0
Fork 0
mirror of https://github.com/msberends/AMR.git synced 2025-07-10 00:23:03 +02:00
Code Releases Activity

(v1.3.0.9001) website update

Browse Source
This commit is contained in:
dr. M.S. (Matthijs) Berends
2020-08-10 12:46:03 +02:00
parent 0d9602a6a9
commit 7d16bec21f
79 changed files with 2221 additions and 1974 deletions
Download Patch File Download Diff File Expand all files Collapse all files

90
docs/articles/resistance_predict.html
View File

@ -39,7 +39,7 @@
</button>
<span class="navbar-brand">
<a class="navbar-link" href="../index.html">AMR (for R)</a>
<span class="version label label-default" data-toggle="tooltip" data-placement="bottom" title="Latest development version">1.3.0</span>
<span class="version label label-default" data-toggle="tooltip" data-placement="bottom" title="Latest development version">1.3.0.9001</span>
</span>
</div>
@ -186,7 +186,7 @@
<h1 data-toc-skip>How to predict antimicrobial resistance</h1>
<h4 class="author">Matthijs S. Berends</h4>
<h4 class="date">30 July 2020</h4>
<h4 class="date">10 August 2020</h4>
<small class="dont-index">Source: <a href="https://github.com/msberends/AMR/blob/master/vignettes/resistance_predict.Rmd"><code>vignettes/resistance_predict.Rmd</code></a></small>
<div class="hidden name"><code>resistance_predict.Rmd</code></div>
@ -200,35 +200,38 @@
<a href="#needed-r-packages" class="anchor"></a>Needed R packages</h2>
<p>As with many uses in R, we need some additional packages for AMR analysis. Our package works closely together with the <a href="https://www.tidyverse.org">tidyverse packages</a> <a href="https://dplyr.tidyverse.org/"><code>dplyr</code></a> and <a href="https://ggplot2.tidyverse.org"><code>ggplot2</code></a> by Dr Hadley Wickham. The tidyverse tremendously improves the way we conduct data science - it allows for a very natural way of writing syntaxes and creating beautiful plots in R.</p>
<p>Our <code>AMR</code> package depends on these packages and even extends their use and functions.</p>
<div class="sourceCode" id="cb1"><html><body><pre class="r"><span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="no">dplyr</span>)
<span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="no">ggplot2</span>)
<span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="no">AMR</span>)
<div class="sourceCode" id="cb1"><pre class="downlit">
<span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="kw"><a href="https://dplyr.tidyverse.org">dplyr</a></span>)
<span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="kw"><a href="http://ggplot2.tidyverse.org">ggplot2</a></span>)
<span class="fu"><a href="https://rdrr.io/r/base/library.html">library</a></span>(<span class="kw"><a href="https://msberends.github.io/AMR">AMR</a></span>)
<span class="co"># (if not yet installed, install with:)</span>
<span class="co"># install.packages(c("tidyverse", "AMR"))</span></pre></body></html></div>
<span class="co"># install.packages(c("tidyverse", "AMR"))</span>
</pre></div>
</div>
<div id="prediction-analysis" class="section level2">
<h2 class="hasAnchor">
<a href="#prediction-analysis" class="anchor"></a>Prediction analysis</h2>
<p>Our package contains a function <code><a href="../reference/resistance_predict.html">resistance_predict()</a></code>, which takes the same input as functions for <a href="./AMR.html">other AMR analysis</a>. Based on a date column, it calculates cases per year and uses a regression model to predict antimicrobial resistance.</p>
<p>It is basically as easy as:</p>
<div class="sourceCode" id="cb2"><html><body><pre class="r"># resistance prediction of piperacillin/tazobactam (TZP):
resistance_predict(tbl = example_isolates, col_date = "date", col_ab = "TZP", model = "binomial")
# or:
example_isolates %&gt;%
resistance_predict(col_ab = "TZP",
model "binomial")
# to bind it to object 'predict_TZP' for example:
predict_TZP &lt;- example_isolates %&gt;%
resistance_predict(col_ab = "TZP",
model = "binomial")</pre></body></html></div>
<div class="sourceCode" id="cb2"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb2-1"><a href="#cb2-1"></a><span class="co"># resistance prediction of piperacillin/tazobactam (TZP):</span></span>
<span id="cb2-2"><a href="#cb2-2"></a><span class="kw">resistance_predict</span>(<span class="dt">tbl =</span> example_isolates, <span class="dt">col_date =</span> <span class="st">"date"</span>, <span class="dt">col_ab =</span> <span class="st">"TZP"</span>, <span class="dt">model =</span> <span class="st">"binomial"</span>)</span>
<span id="cb2-3"><a href="#cb2-3"></a></span>
<span id="cb2-4"><a href="#cb2-4"></a><span class="co"># or:</span></span>
<span id="cb2-5"><a href="#cb2-5"></a>example_isolates <span class="op">%&gt;%</span><span class="st"> </span></span>
<span id="cb2-6"><a href="#cb2-6"></a><span class="st"> </span><span class="kw">resistance_predict</span>(<span class="dt">col_ab =</span> <span class="st">"TZP"</span>,</span>
<span id="cb2-7"><a href="#cb2-7"></a> model <span class="st">"binomial"</span>)</span>
<span id="cb2-8"><a href="#cb2-8"></a></span>
<span id="cb2-9"><a href="#cb2-9"></a><span class="co"># to bind it to object 'predict_TZP' for example:</span></span>
<span id="cb2-10"><a href="#cb2-10"></a>predict_TZP &lt;-<span class="st"> </span>example_isolates <span class="op">%&gt;%</span><span class="st"> </span></span>
<span id="cb2-11"><a href="#cb2-11"></a><span class="st"> </span><span class="kw">resistance_predict</span>(<span class="dt">col_ab =</span> <span class="st">"TZP"</span>,</span>
<span id="cb2-12"><a href="#cb2-12"></a> <span class="dt">model =</span> <span class="st">"binomial"</span>)</span></code></pre></div>
<p>The function will look for a date column itself if <code>col_date</code> is not set.</p>
<p>When running any of these commands, a summary of the regression model will be printed unless using <code><a href="../reference/resistance_predict.html">resistance_predict(..., info = FALSE)</a></code>.</p>
<pre><code># NOTE: Using column `date` as input for `col_date`.</code></pre>
<p>This text is only a printed summary - the actual result (output) of the function is a <code>data.frame</code> containing for each year: the number of observations, the actual observed resistance, the estimated resistance and the standard error below and above the estimation:</p>
<div class="sourceCode" id="cb4"><html><body><pre class="r"><span class="no">predict_TZP</span>
<div class="sourceCode" id="cb4"><pre class="downlit">
<span class="kw">predict_TZP</span>
<span class="co"># year value se_min se_max observations observed estimated</span>
<span class="co"># 1 2002 0.20000000 NA NA 15 0.20000000 0.05616378</span>
<span class="co"># 2 2003 0.06250000 NA NA 32 0.06250000 0.06163839</span>
@ -258,27 +261,36 @@ predict_TZP &lt;- example_isolates %&gt;%
<span class="co"># 26 2027 0.41315710 0.3244399 0.5018743 NA NA 0.41315710</span>
<span class="co"># 27 2028 0.43730688 0.3418075 0.5328063 NA NA 0.43730688</span>
<span class="co"># 28 2029 0.46175755 0.3597639 0.5637512 NA NA 0.46175755</span>
<span class="co"># 29 2030 0.48639359 0.3782932 0.5944939 NA NA 0.48639359</span></pre></body></html></div>
<span class="co"># 29 2030 0.48639359 0.3782932 0.5944939 NA NA 0.48639359</span>
</pre></div>
<p>The function <code>plot</code> is available in base R, and can be extended by other packages to depend the output based on the type of input. We extended its function to cope with resistance predictions:</p>
<div class="sourceCode" id="cb5"><html><body><pre class="r"><span class="fu"><a href="https://rdrr.io/r/base/plot.html">plot</a></span>(<span class="no">predict_TZP</span>)</pre></body></html></div>
<div class="sourceCode" id="cb5"><pre class="downlit">
<span class="fu"><a href="https://rdrr.io/r/graphics/plot.default.html">plot</a></span>(<span class="kw">predict_TZP</span>)
</pre></div>
<p><img src="resistance_predict_files/figure-html/unnamed-chunk-4-1.png" width="720"></p>
<p>This is the fastest way to plot the result. It automatically adds the right axes, error bars, titles, number of available observations and type of model.</p>
<p>We also support the <code>ggplot2</code> package with our custom function <code><a href="../reference/resistance_predict.html">ggplot_rsi_predict()</a></code> to create more appealing plots:</p>
<div class="sourceCode" id="cb6"><html><body><pre class="r"><span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>(<span class="no">predict_TZP</span>)</pre></body></html></div>
<div class="sourceCode" id="cb6"><pre class="downlit">
<span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>(<span class="kw">predict_TZP</span>)
</pre></div>
<p><img src="resistance_predict_files/figure-html/unnamed-chunk-5-1.png" width="720"></p>
<div class="sourceCode" id="cb7"><html><body><pre class="r">
<div class="sourceCode" id="cb7"><pre class="downlit">
<span class="co"># choose for error bars instead of a ribbon</span>
<span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>(<span class="no">predict_TZP</span>, <span class="kw">ribbon</span> <span class="kw">=</span> <span class="fl">FALSE</span>)</pre></body></html></div>
<span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>(<span class="kw">predict_TZP</span>, ribbon = <span class="fl">FALSE</span>)
</pre></div>
<p><img src="resistance_predict_files/figure-html/unnamed-chunk-5-2.png" width="720"></p>
<div id="choosing-the-right-model" class="section level3">
<h3 class="hasAnchor">
<a href="#choosing-the-right-model" class="anchor"></a>Choosing the right model</h3>
<p>Resistance is not easily predicted; if we look at vancomycin resistance in Gram-positive bacteria, the spread (i.e. standard error) is enormous:</p>
<div class="sourceCode" id="cb8"><html><body><pre class="r"><span class="no">example_isolates</span> <span class="kw">%&gt;%</span>
<span class="fu"><a href="https://dplyr.tidyverse.org/reference/filter.html">filter</a></span>(<span class="fu"><a href="../reference/mo_property.html">mo_gramstain</a></span>(<span class="no">mo</span>, <span class="kw">language</span> <span class="kw">=</span> <span class="kw">NULL</span>) <span class="kw">==</span> <span class="st">"Gram-positive"</span>) <span class="kw">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">resistance_predict</a></span>(<span class="kw">col_ab</span> <span class="kw">=</span> <span class="st">"VAN"</span>, <span class="kw">year_min</span> <span class="kw">=</span> <span class="fl">2010</span>, <span class="kw">info</span> <span class="kw">=</span> <span class="fl">FALSE</span>, <span class="kw">model</span> <span class="kw">=</span> <span class="st">"binomial"</span>) <span class="kw">%&gt;%</span>
<div class="sourceCode" id="cb8"><pre class="downlit">
<span class="kw">example_isolates</span> <span class="op">%&gt;%</span>
<span class="fu"><a href="https://dplyr.tidyverse.org/reference/filter.html">filter</a></span>(<span class="fu"><a href="../reference/mo_property.html">mo_gramstain</a></span>(<span class="kw">mo</span>, language = <span class="kw">NULL</span>) <span class="op">==</span> <span class="st">"Gram-positive"</span>) <span class="op">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">resistance_predict</a></span>(col_ab = <span class="st">"VAN"</span>, year_min = <span class="fl">2010</span>, info = <span class="fl">FALSE</span>, model = <span class="st">"binomial"</span>) <span class="op">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>()
<span class="co"># NOTE: Using column `date` as input for `col_date`.</span></pre></body></html></div>
<span class="co"># NOTE: Using column `date` as input for `col_date`.</span>
</pre></div>
<p><img src="resistance_predict_files/figure-html/unnamed-chunk-6-1.png" width="720"></p>
<p>Vancomycin resistance could be 100% in ten years, but might also stay around 0%.</p>
<p>You can define the model with the <code>model</code> parameter. The model chosen above is a generalised linear regression model using a binomial distribution, assuming that a period of zero resistance was followed by a period of increasing resistance leading slowly to more and more resistance.</p>
@ -319,25 +331,29 @@ predict_TZP &lt;- example_isolates %&gt;%
</tbody>
</table>
<p>For the vancomycin resistance in Gram-positive bacteria, a linear model might be more appropriate since no binomial distribution is to be expected based on the observed years:</p>
<div class="sourceCode" id="cb9"><html><body><pre class="r"><span class="no">example_isolates</span> <span class="kw">%&gt;%</span>
<span class="fu"><a href="https://dplyr.tidyverse.org/reference/filter.html">filter</a></span>(<span class="fu"><a href="../reference/mo_property.html">mo_gramstain</a></span>(<span class="no">mo</span>, <span class="kw">language</span> <span class="kw">=</span> <span class="kw">NULL</span>) <span class="kw">==</span> <span class="st">"Gram-positive"</span>) <span class="kw">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">resistance_predict</a></span>(<span class="kw">col_ab</span> <span class="kw">=</span> <span class="st">"VAN"</span>, <span class="kw">year_min</span> <span class="kw">=</span> <span class="fl">2010</span>, <span class="kw">info</span> <span class="kw">=</span> <span class="fl">FALSE</span>, <span class="kw">model</span> <span class="kw">=</span> <span class="st">"linear"</span>) <span class="kw">%&gt;%</span>
<div class="sourceCode" id="cb9"><pre class="downlit">
<span class="kw">example_isolates</span> <span class="op">%&gt;%</span>
<span class="fu"><a href="https://dplyr.tidyverse.org/reference/filter.html">filter</a></span>(<span class="fu"><a href="../reference/mo_property.html">mo_gramstain</a></span>(<span class="kw">mo</span>, language = <span class="kw">NULL</span>) <span class="op">==</span> <span class="st">"Gram-positive"</span>) <span class="op">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">resistance_predict</a></span>(col_ab = <span class="st">"VAN"</span>, year_min = <span class="fl">2010</span>, info = <span class="fl">FALSE</span>, model = <span class="st">"linear"</span>) <span class="op">%&gt;%</span>
<span class="fu"><a href="../reference/resistance_predict.html">ggplot_rsi_predict</a></span>()
<span class="co"># NOTE: Using column `date` as input for `col_date`.</span></pre></body></html></div>
<span class="co"># NOTE: Using column `date` as input for `col_date`.</span>
</pre></div>
<p><img src="resistance_predict_files/figure-html/unnamed-chunk-7-1.png" width="720"></p>
<p>This seems more likely, doesn’t it?</p>
<p>The model itself is also available from the object, as an <code>attribute</code>:</p>
<div class="sourceCode" id="cb10"><html><body><pre class="r"><span class="no">model</span> <span class="kw">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/attributes.html">attributes</a></span>(<span class="no">predict_TZP</span>)$<span class="no">model</span>
<div class="sourceCode" id="cb10"><pre class="downlit">
<span class="kw">model</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/attributes.html">attributes</a></span>(<span class="kw">predict_TZP</span>)<span class="op">$</span><span class="kw">model</span>
<span class="fu"><a href="https://rdrr.io/r/base/summary.html">summary</a></span>(<span class="no">model</span>)$<span class="no">family</span>
<span class="fu"><a href="https://rdrr.io/r/base/summary.html">summary</a></span>(<span class="kw">model</span>)<span class="op">$</span><span class="kw">family</span>
<span class="co"># </span>
<span class="co"># Family: binomial </span>
<span class="co"># Link function: logit</span>
<span class="fu"><a href="https://rdrr.io/r/base/summary.html">summary</a></span>(<span class="no">model</span>)$<span class="no">coefficients</span>
<span class="fu"><a href="https://rdrr.io/r/base/summary.html">summary</a></span>(<span class="kw">model</span>)<span class="op">$</span><span class="kw">coefficients</span>
<span class="co"># Estimate Std. Error z value Pr(&gt;|z|)</span>
<span class="co"># (Intercept) -200.67944891 46.17315349 -4.346237 1.384932e-05</span>
<span class="co"># year 0.09883005 0.02295317 4.305725 1.664395e-05</span></pre></body></html></div>
<span class="co"># year 0.09883005 0.02295317 4.305725 1.664395e-05</span>
</pre></div>
</div>
</div>
</div>
@ -357,7 +373,7 @@ predict_TZP &lt;- example_isolates %&gt;%
</div>
<div class="pkgdown">
<p>Site built with <a href="https://pkgdown.r-lib.org/">pkgdown</a> 1.5.1.</p>
<p>Site built with <a href="https://pkgdown.r-lib.org/">pkgdown</a> 1.5.1.9000.</p>
</div>
</footer>