<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Allan victor on Medium]]></title>
        <description><![CDATA[Stories by Allan victor on Medium]]></description>
        <link>https://medium.com/@victorallan?source=rss-f740847e5509------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*m2D1JuhpGpGPaz_hTmrvig.jpeg</url>
            <title>Stories by Allan victor on Medium</title>
            <link>https://medium.com/@victorallan?source=rss-f740847e5509------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 00:08:46 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@victorallan/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Lets Build a Neural Network in Excel and Python]]></title>
            <link>https://medium.com/@victorallan/lets-build-a-neural-network-in-excel-and-python-26d2e1aa02d9?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/26d2e1aa02d9</guid>
            <category><![CDATA[neural-networks]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[training]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Fri, 14 Mar 2025 09:21:15 GMT</pubDate>
            <atom:updated>2025-03-14T09:26:37.619Z</atom:updated>
            <content:encoded><![CDATA[<p>Back in the day, I used to scratch my head trying to wrap my mind around neural networks and all their fancy formulas. So, here’s the deal: I’m gonna show you how to whip up a basic neural network right in Excel. Trust me, it’s a game-changer for getting a real feel for how these things tick. No more abstract mumbo-jumbo — we’re talking hands-on, see-it-to-believe-it stuff.</p><p><em>Promise : I don&#39;t confuse you with equations and complex math etc.</em></p><p><a href="https://docs.google.com/spreadsheets/d/1g7kmS-YRy5L7WT05au4EcuAzCkhp9DimONOgilclDUo/edit?usp=sharing"><strong>Access Excel/Googlesheets here.</strong></a></p><blockquote>Special Credits to the YouTube channel —<a href="https://youtube.com/@conceptsilluminated?si=MRJzAhojvTpiUXYw"> Concepts Illuminated </a>— Great content and Knowledge regarding NN.</blockquote><h3>Understanding the Basics: Weights and Bias in Neural Networks (NN)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/731/1*Ky1_KerqYGwWVeBWIWX1Xg.jpeg" /><figcaption>NN — Simple architecture — Created by the Author</figcaption></figure><p>In the diagram above, you see a simplified representation of how a neural network processes inputs using weights and bias. Let’s break it down:</p><p><strong>Inputs</strong>: These are the raw numbers fed into the network. In our diagram, we have two sets of inputs, labeled Input 1 and Input 2. Each input is essentially a series of numbers, which might represent features like pixel values in an image or characteristics of a data point.</p><p><strong>Weights</strong>: Weights are crucial in determining the influence of each input on the final output. In the diagram, each input is connected to the output via a weight (Weight 1 and Weight 2). These weights are essentially multipliers that scale the input values, indicating how much importance each input has in predicting the output. For instance, if Weight 1 is larger than Weight 2, Input 1 will have a greater impact on the output compared to Input 2.</p><p><strong>Bias</strong>: Bias acts as a constant added to the weighted sum of inputs. It helps the network make predictions even when all inputs are zero, providing a baseline or threshold for activation. In the diagram, the bias is added to the product of weights and inputs, adjusting the final output value. This addition allows the network to fit the data more flexibly, shifting the activation function left or right as needed.</p><p><strong>Output</strong>: The output is the result of the combined effect of inputs, weights, and bias. It’s the prediction or decision the network makes based on the given inputs. The calculation is straightforward: multiply each input by its respective weight, sum these products, and add the bias. This final value is then passed through an activation function (not shown in the diagram), which determines whether the neuron “fires” or activates.</p><p>By visualizing weights and bias this way, you can better understand how neural networks transform inputs into meaningful outputs. This foundational knowledge will be invaluable as we dive into building a basic neural network in Excel and Python, allowing you to see firsthand how adjustments to weights and bias can change the network’s behavior and accuracy.</p><p>Ready to get hands-on and build your first neural network? Let’s dive into the practical steps using Excel and Python, where theory meets practice!</p><h3>Understanding the Neural Network Setup for AND Gate</h3><p><strong>THE AND GATE:</strong></p><p>The AND function is a basic logical operation that outputs true (1) only when both inputs are true (1), according to the following rules:</p><ul><li>Input 1 and Input 2 are both 0: Output is 0</li><li>Input 1 is 0 and Input 2 is 1: Output is 0</li><li>Input 1 is 1 and Input 2 is 0: Output is 0</li><li>Input 1 and Input 2 are both 1: Output is 1</li></ul><p><strong>THE NN SETUP:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pyrSHmhj35EsdOxLfvry0A.png" /></figure><p><strong>Inputs:</strong></p><ul><li>Input 1 (x_1) and Input 2 (x_2): These are the binary inputs for the AND gate, represented as a series of 0s and 1s. The network is trained to output 1 only when both inputs are 1.</li></ul><p><strong>Weights and Bias:</strong></p><ul><li>Weights (w_1, w_2): Initially set to 1. These weights determine the influence of each input on the output. In a more complex network, these would be adjusted during training to minimize error.</li><li>Bias (b): Also set to 1. The bias allows the network to output values greater than zero even when inputs are zero, providing a baseline activation.</li></ul><p><strong>Weighted Sum (z_3):</strong></p><ul><li>The weighted sum ( z ) is calculated by the Excel formula —</li></ul><p><strong>=(E10*B10)+(F10*C10)+A10</strong></p><ul><li>E10 and F10 are the input values.</li><li>B10 and C10 are the weights for those inputs.</li><li>A10 is the bias added to the sum.</li></ul><p>This sum represents the combined influence of inputs and bias before applying the activation function.</p><p><strong>Output (a_3):</strong></p><ul><li>The output ( a ) is computed using the sigmoid activation function. This function maps or scales the weighted sum to a value between 0 and 1, making it suitable for binary classification tasks like the AND gate.</li></ul><p><strong>Cost :</strong></p><p>Subtract predicted output from the actual output and square the values. The cost should be as low as possible. The complex explanation down below</p><ul><li>The cost ( C ) is calculated using the formula ( C = (y — a)² ), where ( y ) is the desired output. This squared error measure indicates how far the network’s prediction is from the actual output, guiding adjustments during training.</li></ul><p><strong>Activation Chart:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/429/1*XyJbpqpwyvgz1Gg-y-aEHg.png" /></figure><ul><li>The activation chart visualizes the output of the sigmoid function across a range of input values. It represents all the probabilities. The basic distribution for AND function is 1 for 1,1 input and 0 for all others. Hence in the chart, there is bright blue color represent higher value at 1,1 and white at lower values representing 0 at 0,0. This helps illustrate how the function maps weighted sums to activations between 0 and 1.</li></ul><h3>Explanation of the Process:</h3><ol><li>Input Processing: Each pair of inputs is multiplied by its respective weight, summed, and added to the bias to produce the weighted sum.</li><li>Activation: The weighted sum is passed through the sigmoid function, which compresses the value into a range between 0 and 1, simulating the behavior of a neuron “firing.”</li><li>Cost Calculation: The output is compared to the desired output (from the AND gate), and the cost is calculated to quantify the prediction error.</li><li>Learning: In a complete neural network setup, the weights and bias would be adjusted iteratively (in each cycle) using methods like gradient descent to minimize the cost, improving the network’s ability to predict the correct output.</li></ol><p>This Excel model provides a hands-on approach to understanding the fundamental operations of a neural network, illustrating how inputs are transformed into outputs through weighted sums and activation functions.</p><p>Implementing in Python with the output,</p><pre>import numpy as np<br>import matplotlib.pyplot as plt<br><br># Sigmoid function<br>def sigmoid(x):<br>    return 1 / (1 + np.exp(-x))<br><br># Derivative of the sigmoid function<br>def sigmoid_derivative(x):<br>    return sigmoid(x) * (1 - sigmoid(x))<br><br># Training data for AND gate<br>inputs = np.array([[0, 0],<br>                   [0, 1],<br>                   [1, 0],<br>                   [1, 1]])<br><br># Expected output for AND gate<br>expected_output = np.array([[0], [0], [0], [1]])<br><br># Initialize weights and bias<br>weights = np.random.rand(2, 1)  # Two inputs, one output<br>bias = np.random.rand(1)<br><br># Learning rate<br>learning_rate = 0.1<br><br># Number of epochs<br>epochs = 10000<br><br># Lists to store the error for plotting<br>errors = []<br><br># Training the neural network<br>for epoch in range(epochs):<br>    # Forward pass<br>    weighted_sum = np.dot(inputs, weights) + bias<br>    output = sigmoid(weighted_sum)<br><br>    # Calculate error<br>    error = expected_output - output<br>    errors.append(np.mean(np.abs(error)))<br><br>    # Backward pass<br>    adjustments = error * sigmoid_derivative(weighted_sum)<br>    weights += np.dot(inputs.T, adjustments) * learning_rate<br>    bias += np.sum(adjustments) * learning_rate<br><br># Output the trained weights and bias<br>print(&quot;Trained weights:&quot;, weights)<br>print(&quot;Trained bias:&quot;, bias)<br><br># Plot the error over epochs<br>plt.figure(figsize=(12, 5))<br>plt.subplot(1, 2, 1)<br>plt.plot(errors)<br>plt.title(&#39;Error Over Epochs&#39;)<br>plt.xlabel(&#39;Epoch&#39;)<br>plt.ylabel(&#39;Mean Absolute Error&#39;)<br><br># Plot the decision boundary<br>plt.subplot(1, 2, 2)<br>plt.title(&#39;Decision Boundary&#39;)<br><br># Generate a grid of points<br>x_values = np.linspace(0, 1, 100)<br>y_values = np.linspace(0, 1, 100)<br>xx, yy = np.meshgrid(x_values, y_values)<br>grid = np.c_[xx.ravel(), yy.ravel()]<br><br># Compute the output for each point in the grid<br>grid_output = sigmoid(np.dot(grid, weights) + bias)<br>grid_output = grid_output.reshape(xx.shape)<br><br># Plot the decision boundary<br>plt.contourf(xx, yy, grid_output, levels=[0, 0.5, 1], cmap=&#39;coolwarm&#39;, alpha=0.3)<br>plt.scatter(inputs[:, 0], inputs[:, 1], c=expected_output.ravel(), cmap=&#39;coolwarm&#39;, edgecolors=&#39;k&#39;)<br>plt.xlabel(&#39;Input 1&#39;)<br>plt.ylabel(&#39;Input 2&#39;)<br><br>plt.tight_layout()<br>plt.show()<br><br># Test the neural network<br>test_input = np.array([1, 1])<br>test_output = sigmoid(np.dot(test_input, weights) + bias)<br>print(&quot;Output for [1, 1]:&quot;, test_output)</pre><pre>Trained weights: [[5.47786552]<br> [5.47786551]]<br>Trained bias: [-8.30949732]<br>Output for [1, 1]: [0.93377848]</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xK9kAGAGuzf9OlRT7zVyEA.png" /></figure><h3>Thank You for Exploring Neural Networks with Me!</h3><p>I hope this journey into the world of neural networks has illuminated the complexities and simplified the fundamentals for you. By combining Excel and Python, you’ve taken a significant step in understanding how these powerful tools transform inputs into meaningful outputs.</p><h4>Stay Connected:</h4><ul><li>Follow me on <a href="http://www.linkedin.com/in/dr-allan-victor-6919b4152">LinkedIn </a>and <a href="https://medium.com/@victorallan">Medium </a>for more insights and tutorials.</li></ul><h4>Special Thanks:</h4><ul><li>A heartfelt thank you to the YouTube channel “Concepts Illuminated” for their inspiring content and invaluable knowledge.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=26d2e1aa02d9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Create Stunning Chord Diagrams in Minutes: An Open-Source R Shiny Solution]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/create-stunning-chord-diagrams-in-minutes-an-open-source-r-shiny-solution-f98065f2b723?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1919/1*PYOIx4yzV-qhSdUG08vV4g.png" width="1919"></a></p><p class="medium-feed-snippet">Want to create beautiful chord diagrams without the hassle? As a data scientist, I know the pain of spending hours on visualization code&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/create-stunning-chord-diagrams-in-minutes-an-open-source-r-shiny-solution-f98065f2b723?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/create-stunning-chord-diagrams-in-minutes-an-open-source-r-shiny-solution-f98065f2b723?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/f98065f2b723</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[visualization]]></category>
            <category><![CDATA[biotechnology]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Wed, 02 Oct 2024 13:42:48 GMT</pubDate>
            <atom:updated>2024-10-02T13:42:48.489Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[No-Code Spider Chart Generator in plotly: Easily Visualize Your AI and ML Models]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/no-code-spider-chart-generator-in-plotly-easily-visualize-your-ai-and-ml-models-ed64c93affed?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1919/1*-tnYdD8IXT8Q1toPTmCr2A.png" width="1919"></a></p><p class="medium-feed-snippet">This generator can be used for any data not exclusive to AI and ML models</p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/no-code-spider-chart-generator-in-plotly-easily-visualize-your-ai-and-ml-models-ed64c93affed?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/no-code-spider-chart-generator-in-plotly-easily-visualize-your-ai-and-ml-models-ed64c93affed?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/ed64c93affed</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Sat, 28 Sep 2024 14:53:47 GMT</pubDate>
            <atom:updated>2024-09-28T14:53:47.107Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Deinfluencing: Why Your 9-to-5 Jobs Isn’t the ‘Rat Race’ as Influencers Claim]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/deinfluencing-why-your-9-to-5-jobs-isnt-the-rat-race-as-influencers-claim-1defddfe5e3c?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/2600/1*GooYXQZ1bAEa-GuSn_NJGA.jpeg" width="5304"></a></p><p class="medium-feed-snippet">In an era where social media influencers constantly bombard us with images of lavish lifestyles and promises of &#x201C;easy money,&#x201D; it&#x2019;s time to&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/deinfluencing-why-your-9-to-5-jobs-isnt-the-rat-race-as-influencers-claim-1defddfe5e3c?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/deinfluencing-why-your-9-to-5-jobs-isnt-the-rat-race-as-influencers-claim-1defddfe5e3c?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/1defddfe5e3c</guid>
            <category><![CDATA[money]]></category>
            <category><![CDATA[personal-development]]></category>
            <category><![CDATA[self-improvement]]></category>
            <category><![CDATA[work-life-balance]]></category>
            <category><![CDATA[personal-finance]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Thu, 26 Sep 2024 06:57:28 GMT</pubDate>
            <atom:updated>2024-09-26T06:57:28.936Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Bridging the Gap: Converting SSR Data to HapMap for GWAS in TASSEL]]></title>
            <link>https://medium.com/@victorallan/bridging-the-gap-converting-ssr-data-to-hapmap-for-gwas-in-tassel-69f20554e2fe?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/69f20554e2fe</guid>
            <category><![CDATA[bioinformatics]]></category>
            <category><![CDATA[tassels]]></category>
            <category><![CDATA[association]]></category>
            <category><![CDATA[gwas]]></category>
            <category><![CDATA[genomics]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Mon, 23 Sep 2024 13:23:14 GMT</pubDate>
            <atom:updated>2024-09-27T15:40:35.581Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*in6CKlWbnsQgwI2C.gif" /><figcaption>Source:<a href="https://avikarn.com/2019-07-22-GWAS/">https://avikarn.com/2019-07-22-GWAS/</a></figcaption></figure><p>In the ever-evolving landscape of genomics research, staying current with the latest tools and data formats is crucial. Recently, TASSEL (Trait Analysis by aSSociation, Evolution and Linkage), a popular software for Genome-Wide Association Studies (GWAS), made a significant change by removing support for Simple Sequence Repeat (SSR) data. This decision has left many researchers, particularly those working with non-model organisms or historical datasets, in a challenging position. But fear not! We’ve developed a solution to bridge this gap and ensure your valuable SSR data can still be used for GWAS in TASSEL.</p><h3>The Challenge: SSR Data in a SNP World</h3><p>Simple Sequence Repeats, also known as microsatellites, have been a staple in genetic studies for decades. They’re particularly useful in populations with low genetic diversity or when working with organisms that lack extensive genomic resources. However, the trend in genomics has been moving towards Single Nucleotide Polymorphisms (SNPs) due to their abundance and ease of high-throughput genotyping.</p><p>TASSEL’s decision to remove SSR support reflects this shift, but it doesn’t mean your SSR data is obsolete. The key is to convert your SSR data into a format that TASSEL can understand — enter the HapMap format.</p><h3>The Solution: SSR to HapMap Converter</h3><p>We’ve developed a user-friendly Shiny app that converts your SSR data into the HapMap format, making it compatible with TASSEL for GWAS analysis. Here’s how it works:</p><ol><li><strong>Data Input</strong>: Upload your SSR data in STRUCTURE format (.csv) — <a href="https://drive.google.com/file/d/1cVdUuyrrjlwatAhZnhe8pUkpeULPMSjU/view?usp=sharing">download data</a>, and a metadata file — <a href="https://drive.google.com/file/d/1Ey_at_p8edMnHyTiWuV2NVchHT8NXWMJ/view?usp=sharing">download data</a>.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pOU2Sz2ywYRiqrTP5L_cUw.png" /><figcaption>SSR data in STRUCTURE format (.csv)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JC7mfSokuzCwH5wyMKf29g.png" /><figcaption>metadata file format</figcaption></figure><p><strong>2. Data Processing</strong>: The app processes your SSR data, converting the alleles to a SNP-like format based on frequency.</p><p><strong>3. HapMap Output</strong>: The processed data is formatted into HapMap format, ready for download and use in TASSEL.</p><h3>How to Use the Converter</h3><ol><li>Just copy paste the code in R-Studio and select all and Click RUN — Launch the Shiny app (code provided below).</li><li>Upload your SSR data file (CSV format) and metadata file.</li><li>The app will process your data and display it in a table for review.</li><li>You can view a summary of your marker data, including the number of markers and samples.</li><li>Download the converted data in HapMap format.</li></ol><pre># Load required libraries - Install the required packages if not installed <br>library(shiny)<br>library(DT)<br><br># Define UI<br>ui &lt;- fluidPage(<br>  titlePanel(&quot;SSR to HapMap Converter&quot;),<br>  <br>  sidebarLayout(<br>    sidebarPanel(<br>      fileInput(&quot;ssrFile&quot;, &quot;Upload SSR Data (CSV)&quot;, accept = &quot;.csv&quot;),<br>      fileInput(&quot;metaFile&quot;, &quot;Upload Metadata (CSV)&quot;, accept = &quot;.csv&quot;),<br>      downloadButton(&quot;downloadHapmap&quot;, &quot;Download HapMap File&quot;),<br>      width = 3<br>    ),<br>    <br>    mainPanel(<br>      tabsetPanel(<br>        tabPanel(&quot;Processed Data&quot;, DTOutput(&quot;processedData&quot;)),<br>        tabPanel(&quot;Marker Summary&quot;, verbatimTextOutput(&quot;markerSummary&quot;))<br>      )<br>    )<br>  )<br>)<br><br># Define server logic<br>server &lt;- function(input, output, session) {<br>  <br>  # Reactive function to process the data<br>  processData &lt;- reactive({<br>    req(input$ssrFile, input$metaFile)<br>    <br>    # Read SSR data<br>    data &lt;- read.csv(input$ssrFile$datapath, na.strings = &quot;-9&quot;)<br>    <br>    # Read metadata<br>    meta_data &lt;- read.csv(input$metaFile$datapath)<br>    <br>    # Process SSR data<br>    geno &lt;- unique(data[,1])<br>    mark &lt;- colnames(data)[-1]<br>    code &lt;- c(&quot;A&quot;, &quot;C&quot;, &quot;G&quot;, &quot;T&quot;, &quot;+&quot;, &quot;-&quot;)<br>    res &lt;- data.frame(GENO=geno)<br>    <br>    for (i in 2:(ncol(data)-1)) {<br>      x &lt;- data[,i]<br>      u &lt;- table(x)<br>      mark_name &lt;- colnames(data)[i]<br>      for (j in 1:length(u)) {<br>        x[which(x == as.numeric(names(which.max(u))))] &lt;- code[j]<br>        u &lt;- u[-which.max(u)]<br>      }<br>      x1 &lt;- x[seq(1, length(x), 2)]<br>      x2 &lt;- x[seq(2, length(x), 2)]<br>      x &lt;- paste(x1, x2, sep = &quot;&quot;)<br>      x[grep(&quot;NA&quot;, x, fixed = TRUE)] &lt;- &quot;NN&quot;<br>      x[grep(&quot;M&quot;, x, fixed = TRUE)] &lt;- &quot;NN&quot;<br>      <br>      res[mark_name] &lt;- x<br>    }<br>    <br>    rownames(res) &lt;- geno<br>    res &lt;- res[,-1]<br>    res &lt;- t(res)<br>    <br>    # Merge with metadata<br>    temp &lt;- intersect(rownames(res), meta_data$rs)<br>    meta_data_f &lt;- meta_data[na.omit(match(temp, meta_data$rs)),]<br>    res_f &lt;- res[na.omit(match(temp, rownames(res))),]<br>    <br>    if (!all(rownames(res_f) == meta_data_f$rs)) {<br>      stop(&quot;Mismatch between processed data and metadata&quot;)<br>    }<br>    <br>    res_f &lt;- cbind(meta_data_f, res_f)<br>    return(res_f)<br>  })<br>  <br>  # Output processed data<br>  output$processedData &lt;- renderDT({<br>    datatable(processData())<br>  })<br>  <br>  # Output marker summary<br>  output$markerSummary &lt;- renderPrint({<br>    data &lt;- processData()<br>    cat(&quot;Number of markers:&quot;, nrow(data), &quot;\n&quot;)<br>    cat(&quot;Number of samples:&quot;, ncol(data) - ncol(read.csv(input$metaFile$datapath)), &quot;\n&quot;)<br>    cat(&quot;\nAllele frequency summary:\n&quot;)<br>    print(summary(apply(data[, -(1:ncol(read.csv(input$metaFile$datapath)))], 1, function(x) {<br>      table(unlist(strsplit(x, &quot;&quot;)))<br>    })))<br>  })<br>  <br>  # Download handler for HapMap file<br>  output$downloadHapmap &lt;- downloadHandler(<br>    filename = function() {<br>      paste(&quot;hapmap_&quot;, Sys.Date(), &quot;.hmp&quot;, sep = &quot;&quot;)<br>    },<br>    content = function(file) {<br>      write.table(processData(), file, row.names = FALSE, quote = FALSE, sep = &quot;\t&quot;)<br>    }<br>  )<br>}<br><br># Run the Shiny app<br>shinyApp(ui = ui, server = server)</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qD_WtZ2Jpk-ADcjhVwoQHg.png" /><figcaption>The UI of Shiny app</figcaption></figure><h3>Next Steps: GWAS in TASSEL</h3><p>Once you have your HapMap file, you’re ready to proceed with GWAS in TASSEL. For a comprehensive guide on conducting GWAS using TASSEL, we recommend following the excellent tutorial by Avi Karn: <a href="https://avikarn.com/2019-07-22-GWAS/">GWAS in TASSEL</a>.</p><p>This tutorial covers:</p><ul><li>Data preparation and quality control</li><li>Population structure analysis</li><li>Model selection</li><li>Running GWAS</li><li>Visualizing and interpreting results</li></ul><h3>Why This Matters</h3><ol><li><strong>Data Continuity</strong>: This converter allows researchers to continue using valuable historical SSR datasets in modern genomic analyses.</li><li><strong>Non-Model Organisms</strong>: For species without extensive SNP resources, this tool provides a pathway to leverage SSR data for GWAS.</li><li><strong>Comparative Studies</strong>: Researchers can now more easily combine or compare results from SSR-based studies with SNP-based studies.</li></ol><p>&lt;image: Infographic highlighting the benefits of the converter&gt;</p><h3>Limitations and Considerations</h3><p>While this converter bridges an important gap, it’s crucial to understand its limitations:</p><ul><li>The conversion process involves some data loss, as SSR markers typically have more alleles than can be represented in a bi-allelic SNP format.</li><li>The statistical power of your GWAS may be affected by the conversion, particularly if you have a small number of SSR markers.</li><li>Always interpret your results with these limitations in mind and consider validating key findings with additional methods.</li></ul><h3>Conclusion</h3><p>The SSR to HapMap converter represents a practical solution to a pressing problem in genomics research. By allowing the continued use of SSR data in TASSEL, we hope to support ongoing research in diverse organisms and enable more comprehensive meta-analyses across different marker types.</p><p>We encourage researchers to try out this tool and provide feedback. As the genomics field continues to evolve, it’s through community-driven solutions like this that we can ensure no valuable data is left behind.</p><p>Happy researching, and may your GWAS be ever insightful!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=69f20554e2fe" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Create Stunning Network Graphs Without Code: Leveraging Python and R Packages]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/create-stunning-network-graphs-without-code-leveraging-python-and-r-packages-57df86d8d501?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1919/1*mygwF2H_-QPduP-ZuwcRPQ.png" width="1919"></a></p><p class="medium-feed-snippet">In today&#x2019;s data-driven world, visualizing complex relationships is crucial. Whether you&#x2019;re mapping social networks, analyzing supply&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/create-stunning-network-graphs-without-code-leveraging-python-and-r-packages-57df86d8d501?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/create-stunning-network-graphs-without-code-leveraging-python-and-r-packages-57df86d8d501?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/57df86d8d501</guid>
            <category><![CDATA[vision]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[visualization]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Fri, 20 Sep 2024 17:58:59 GMT</pubDate>
            <atom:updated>2024-09-20T17:58:59.352Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Create Stunning Raincloud Plots in Just Two Clicks with Our Shiny App]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/create-stunning-raincloud-plots-in-just-two-clicks-with-our-shiny-app-64b29e22ee43?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1920/1*cbEd2H2-x5kwgC1rJCzl3Q.png" width="1920"></a></p><p class="medium-feed-snippet">Kindly add the errors and difficulties in the comments, I will try to improve or reply in a day or two.</p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/create-stunning-raincloud-plots-in-just-two-clicks-with-our-shiny-app-64b29e22ee43?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/create-stunning-raincloud-plots-in-just-two-clicks-with-our-shiny-app-64b29e22ee43?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/64b29e22ee43</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[r-programming]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Thu, 19 Sep 2024 16:11:52 GMT</pubDate>
            <atom:updated>2024-09-19T16:39:37.623Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Visualizing Flows with Sankey Diagrams: A Step-by-Step Guide in R plotly and networkD3]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/visualizing-flows-with-sankey-diagrams-a-step-by-step-guide-in-r-plotly-and-networkd3-35b2a39f2710?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/895/1*ewG945cxYofgeFqDW2P8hA.jpeg" width="895"></a></p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/visualizing-flows-with-sankey-diagrams-a-step-by-step-guide-in-r-plotly-and-networkd3-35b2a39f2710?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/visualizing-flows-with-sankey-diagrams-a-step-by-step-guide-in-r-plotly-and-networkd3-35b2a39f2710?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/35b2a39f2710</guid>
            <category><![CDATA[bioinformatics]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Sun, 01 Sep 2024 07:27:12 GMT</pubDate>
            <atom:updated>2024-09-13T01:21:12.207Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Visualizing Relationships with Venn Diagrams in R’s ggVennDiagram Package]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/visualizing-relationships-with-venn-diagrams-in-rs-ggvenndiagram-package-066795e126cb?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1651/1*HFOkTVq3694G6Cj8R0V3lw.jpeg" width="1651"></a></p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/visualizing-relationships-with-venn-diagrams-in-rs-ggvenndiagram-package-066795e126cb?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/visualizing-relationships-with-venn-diagrams-in-rs-ggvenndiagram-package-066795e126cb?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/066795e126cb</guid>
            <category><![CDATA[r]]></category>
            <category><![CDATA[bioinformatics]]></category>
            <category><![CDATA[venn-diagrams]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-visualization]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Mon, 18 Dec 2023 06:02:23 GMT</pubDate>
            <atom:updated>2024-09-13T01:22:14.227Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[PB-GWAS: A shiny app for Streamlined GWAS Analysis in Just 4 Clicks ]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@victorallan/pb-gwas-a-shiny-app-for-streamlined-gwas-analysis-in-just-4-clicks-d45f4a286a21?source=rss-f740847e5509------2"><img src="https://cdn-images-1.medium.com/max/1500/1*5LkHEGdQuTxdsqFvcweMaQ.png" width="1500"></a></p><p class="medium-feed-link"><a href="https://medium.com/@victorallan/pb-gwas-a-shiny-app-for-streamlined-gwas-analysis-in-just-4-clicks-d45f4a286a21?source=rss-f740847e5509------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@victorallan/pb-gwas-a-shiny-app-for-streamlined-gwas-analysis-in-just-4-clicks-d45f4a286a21?source=rss-f740847e5509------2</link>
            <guid isPermaLink="false">https://medium.com/p/d45f4a286a21</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[next-generation]]></category>
            <category><![CDATA[genomics]]></category>
            <category><![CDATA[gwas]]></category>
            <category><![CDATA[bioinformatics]]></category>
            <dc:creator><![CDATA[Allan victor]]></dc:creator>
            <pubDate>Thu, 14 Dec 2023 18:29:37 GMT</pubDate>
            <atom:updated>2024-09-13T01:19:53.015Z</atom:updated>
        </item>
    </channel>
</rss>