{"id":3519,"date":"2023-08-31T11:12:58","date_gmt":"2023-08-31T11:12:58","guid":{"rendered":"https:\/\/neuraldesigner.com\/learning\/qsar-biodegradation\/"},"modified":"2026-08-07T11:44:05","modified_gmt":"2026-08-07T09:44:05","slug":"qsar-biodegradation","status":"publish","type":"learning","link":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/","title":{"rendered":"Predict chemical biodegradability with machine learning"},"content":{"rendered":"<style>\n.nds{--nds-code-bg:#f3f7fa;--nds-code-fg:#12354b;width:100vw;margin-left:calc(50% - 50vw);padding:22px 24px 14px;background:#eee;color:#1b2635;font-family:\"Outfit\",\"Roboto\",Arial,sans-serif}\n.nds *{box-sizing:border-box}.nds-wrap{width:min(100%,1200px);margin:0 auto}.nds a{text-decoration:none;color:#2d799f}\n.nds-executive{margin:0 0 34px;padding:32px;border-radius:20px;background:linear-gradient(135deg,#12354b,#245e80);color:#fff}.nds-executive h2{margin:0 0 12px;color:#fff;font-size:30px}.nds-executive p{font-size:18px;line-height:1.55}\n.nds-kpis{display:grid;grid-template-columns:repeat(4,minmax(0,1fr));gap:14px;margin-top:22px}.nds-kpi{padding:18px;border:1px solid rgba(255,255,255,.18);border-radius:14px;background:rgba(255,255,255,.1)}.nds-kpi strong{display:block;font-size:25px}.nds-kpi span{font-size:13px}\n.nds-actions,.nds-audience,.nds-downloads{display:flex;flex-wrap:wrap;justify-content:center;gap:12px;margin:22px 0}.nds-actions a,.nds-downloads a{padding:12px 20px;border-radius:24px;background:#245e80;color:#fff;font-weight:700}.nds-executive .nds-actions a{background:#fff;color:#12354b}\n.nds-toc{display:flex;flex-wrap:wrap;justify-content:center;gap:10px;margin:0 0 42px;padding:0;list-style:none}.nds-toc a,.nds-audience span{display:block;padding:9px 14px;border-radius:20px;background:#e9f2f8;color:#12354b;font-weight:600}\n.nds-card{margin:0 0 54px;scroll-margin-top:90px}.nds-card h2{margin:0 0 20px;padding-bottom:12px;border-bottom:1px solid #dbe5ec;color:#001233;font-size:24px}.nds-card p,.nds-card li{font-size:16.5px;line-height:1.6}.nds-card img{display:block;width:auto;max-width:min(640px,100%);height:auto;margin:24px auto;border-radius:12px}.nds-card img.nds-architecture{width:min(1000px,100%);max-width:100%}\n.nds-value-grid{display:grid;grid-template-columns:repeat(3,minmax(0,1fr));gap:16px;margin:22px 0}.nds-value{padding:20px;border:1px solid #dce8ef;border-radius:14px;background:#f8fbfd}.nds-note{margin:20px 0;padding:18px 20px;border-left:4px solid #56a1c8;border-radius:0 12px 12px 0;background:#f6fafc}.nds-note--warning{border-left-color:#d79a29;background:#fff9ed}\n.nds-table-scroll{max-width:100%;overflow-x:auto}.nds-card table{width:auto;max-width:100%;margin:24px auto;border-collapse:collapse;background:#fff}.nds-card th,.nds-card td{padding:11px 16px;border-bottom:1px solid #e2e9ee}.nds-card thead th{background:#12354b!important;color:#fff!important}.nds-card tbody th{background:#eaf2f6!important;color:#12354b!important;text-align:left}.nds-card tbody tr:nth-child(even) th{background:#f4f8fa!important}.nds-card tbody td{color:#33424f!important}\n.nds-figure-grid{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));gap:20px}.nds-figure-grid figure,.nds-architecture-figure{margin:0;padding:16px;border:1px solid #dce8ef;border-radius:15px;background:#f8fbfd}.nds-figure-grid img{width:100%;max-width:100%;margin:0 auto 12px}.nds-card figcaption{color:#405361;font-size:14px;line-height:1.45}\n.nds pre,.nds .wp-block-code,.nds pre[class*=\"language-\"]{display:block;max-width:100%;overflow:auto;padding:18px 20px!important;border:1px solid #d5e2e9!important;border-radius:12px!important;background:var(--nds-code-bg)!important;color:var(--nds-code-fg)!important;box-shadow:none!important;text-shadow:none!important;white-space:pre}\n.nds pre code,.nds pre code *,.nds .wp-block-code code,.nds .wp-block-code code *{border:0!important;background:transparent!important;color:var(--nds-code-fg)!important;box-shadow:none!important;text-shadow:none!important;font-family:Consolas,\"Liberation Mono\",monospace!important;font-size:14px;line-height:1.55}\n.nds :not(pre)>code{padding:.08em .32em;border-radius:4px;background:#e6eef3!important;color:#12354b!important}\n.nds .nds-expression{max-width:100%;overflow:auto;margin:22px auto;padding:18px 20px;border:1px solid #d5e2e9;border-radius:12px;background:#fff!important;color:#12354b!important;text-align:center}\n.nds .nds-expression math,.nds .nds-expression mjx-container,.nds .nds-expression .MathJax{background:transparent!important;color:#12354b!important}\n@media(max-width:820px){.nds-kpis{grid-template-columns:repeat(2,minmax(0,1fr))}.nds-value-grid{grid-template-columns:1fr}}\n@media(max-width:620px){.nds{padding:12px 14px}.nds-kpis,.nds-figure-grid{grid-template-columns:1fr}.nds-executive{padding:24px 20px}.nds-card table{font-size:13px}.nds-card th,.nds-card td{padding:9px 8px}}\n<\/style>\n<style>\n.nds-details{margin:18px 0 24px;padding:14px 18px;border:1px solid #dce8ef;border-radius:12px;background:#f8fbfd}\n.nds-details summary{cursor:pointer;color:#12354b;font-weight:700}\n.nds-centered-figure,.nds-centered-roc{margin:24px auto!important;padding:16px;border:1px solid #dce8ef;border-radius:15px;background:#f8fbfd}\n.nds-centered-figure{max-width:820px}.nds-centered-roc{display:block;width:min(620px,100%);max-width:620px}\n.nds-centered-figure img,.nds-centered-roc img{display:block;width:100%;max-width:100%;margin:0 auto 12px}\n.nds-flow{display:grid;grid-template-columns:repeat(6,minmax(0,1fr));gap:10px;margin:24px 0}\n.nds-flow div{position:relative;display:flex;min-height:96px;align-items:center;justify-content:center;padding:14px 9px;border-radius:13px;background:#12354b;color:#fff;text-align:center;font-weight:700}\n.nds-flow div:not(:last-child):after{content:\"\u2192\";position:absolute;right:-15px;top:50%;z-index:2;transform:translateY(-50%);color:#56a1c8;font-size:21px}\n.nds-metrics td:nth-child(2),.nds-confusion td{text-align:right;font-variant-numeric:tabular-nums}\n.nds-deployment-grid{display:grid;grid-template-columns:minmax(0,1.6fr) minmax(260px,1fr);gap:22px;align-items:center}\n.nds-result{padding:26px;border:1px solid #cfe1eb;border-radius:16px;background:#f7fbfd;text-align:center}\n.nds-result span,.nds-result strong{display:block}.nds-result strong{margin:8px 0;color:#12354b;font-size:31px}\n@media(max-width:1000px){.nds-flow{grid-template-columns:repeat(3,1fr)}.nds-flow div:after{display:none}}\n@media(max-width:760px){.nds-flow,.nds-deployment-grid{grid-template-columns:1fr}.nds-flow div:after{display:none}}\n<\/style>\n<div class=\"nds\" data-science-profile=\"environmental-ecological\">\n<div class=\"nds-wrap\">\n<section class=\"nds-executive\">\n<h2>Screen chemical structures for ready biodegradability<\/h2>\n<p>This reproducible QSAR tutorial maps 41 molecular descriptors to an internal ready-biodegradability screening score. On 211 held-out chemicals, the exported model reaches 88.2% accuracy, 0.867 balanced accuracy and an ROC AUC of 0.931. The result is an internal benchmark, not a regulatory prediction.<\/p>\n<div class=\"nds-kpis\">\n<div class=\"nds-kpi\"><strong>0.931<\/strong><span>testing ROC AUC<\/span><\/div>\n<div class=\"nds-kpi\"><strong>88.2%<\/strong><span>testing accuracy<\/span><\/div>\n<div class=\"nds-kpi\"><strong>0.867<\/strong><span>balanced accuracy<\/span><\/div>\n<div class=\"nds-kpi\"><strong>211<\/strong><span>held-out chemicals<\/span><\/div>\n<\/div>\n<div class=\"nds-actions\"><a href=\"#6-validation\">Review scientific validation<\/a><a href=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation.csv\">Download the data<\/a><\/div>\n<\/section>\n<ul class=\"nds-toc\">\n<li><a href=\"#1-scientific-objective\">Scientific objective<\/a><\/li>\n<li><a href=\"#2-data-provenance\">Data and provenance<\/a><\/li>\n<li><a href=\"#3-model\">Model<\/a><\/li>\n<li><a href=\"#4-training\">Training<\/a><\/li>\n<li><a href=\"#5-selection\">Selection<\/a><\/li>\n<li><a href=\"#6-validation\">Validation<\/a><\/li>\n<li><a href=\"#7-inference\">Inference<\/a><\/li>\n<li><a href=\"#8-validity\">Validity<\/a><\/li>\n<\/ul>\n<section id=\"1-scientific-objective\" class=\"nds-card\">\n<h2>1. Scientific objective<\/h2>\n<p>The objective is to demonstrate how a transparent tabular QSAR can prioritize chemicals by their likelihood of meeting the source data set&#8217;s ready-biodegradability label. The model can support method development and pre-screening when descriptor generation, chemical identity, applicability and uncertainty are controlled. It does not replace an OECD 301 study or an expert regulatory assessment.<\/p>\n<div class=\"nds-value-grid\">\n<div class=\"nds-value\"><strong>Early chemical screening<\/strong><\/p>\n<p>Prioritize structures for experimental work before committing laboratory time and material.<\/p>\n<\/div>\n<div class=\"nds-value\"><strong>Transparent operating point<\/strong><\/p>\n<p>Compare the default decision threshold with a sensitivity-oriented screening threshold.<\/p>\n<\/div>\n<div class=\"nds-value\"><strong>Reproducible QSAR asset<\/strong><\/p>\n<p>Inspect the renamed schema, exact split, Neural Designer project and executable Python export.<\/p>\n<\/div>\n<\/div>\n<div class=\"nds-audience\">\n<span>Environmental chemists<\/span><span>Computational toxicologists<\/span><span>Regulatory scientists<\/span><span>Chemical R&amp;D<\/span><span>Sustainability teams<\/span>\n<\/div>\n<div class=\"nds-note\"><strong>Scope.<\/strong> The target reproduces a binary label from a public molecular-descriptor benchmark. The page demonstrates model development and screening logic; it does not establish regulatory validity, an applicability domain or laboratory equivalence.<\/div>\n<\/section>\n<section id=\"2-data-provenance\" class=\"nds-card\">\n<h2>2. Data and provenance<\/h2>\n<p>The updated <a href=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation.csv\"><code>biodegradation.csv<\/code><\/a> contains 1,055 chemicals, 41 numeric molecular descriptors and no missing values. The target contains 356 readily biodegradable and 699 not readily biodegradable records.<\/p>\n<div class=\"nds-table-scroll\">\n<table>\n<thead>\n<tr>\n<th>Subset<\/th>\n<th>Rows<\/th>\n<th>Readily biodegradable<\/th>\n<th>Purpose<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th>Training<\/th>\n<td>633<\/td>\n<td>Model-weighted during fitting<\/td>\n<td>Estimate 42 model parameters<\/td>\n<\/tr>\n<tr>\n<th>Selection<\/th>\n<td>211<\/td>\n<td>Monitored during optimization<\/td>\n<td>Track generalization during training<\/td>\n<\/tr>\n<tr>\n<th>Testing<\/th>\n<td>211<\/td>\n<td>73 positive; 138 negative<\/td>\n<td>Report final internal performance<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<h3>Descriptor schema<\/h3>\n<p>The local CSV preserves the original values and column order but replaces compact source codes with descriptive field names. Neural Designer consumes all 41 fields in this exact order.<\/p>\n<details class=\"nds-details\">\n<summary>View the exact 41-column input contract<\/summary>\n<div class=\"nds-table-scroll\">\n<table>\n<thead>\n<tr>\n<th>Published CSV field<\/th>\n<th>Source descriptor<\/th>\n<th>Family<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th><code>laplace_leading_eigenvalue<\/code><\/th>\n<td><code>SpMax_L<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>barysz_balaban_electronegativity_index<\/code><\/th>\n<td><code>J_Dz(e)<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>heavy_atom_count<\/code><\/th>\n<td><code>nHM<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>nitrogen_nitrogen_pair_frequency_distance_1<\/code><\/th>\n<td><code>F01[N-N]<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_nitrogen_pair_frequency_distance_4<\/code><\/th>\n<td><code>F04[C-N]<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>ssss_carbon_atom_count<\/code><\/th>\n<td><code>NssssC<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>substituted_benzene_sp2_carbon_count<\/code><\/th>\n<td><code>nCb-<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_atom_percentage<\/code><\/th>\n<td><code>C%<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>terminal_primary_sp3_carbon_count<\/code><\/th>\n<td><code>nCp<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>oxygen_atom_count<\/code><\/th>\n<td><code>nO<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_nitrogen_pair_frequency_distance_3<\/code><\/th>\n<td><code>F03[C-N]<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>dss_carbon_e_state_sum<\/code><\/th>\n<td><code>SdssC<\/code><\/td>\n<td>Electronic\/E-state<\/td>\n<\/tr>\n<tr>\n<th><code>burden_mass_hyper_wiener_log_index<\/code><\/th>\n<td><code>HyWi_B(m)<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>lopping_centric_index<\/code><\/th>\n<td><code>LOC<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>laplace_spectral_moment_order_6<\/code><\/th>\n<td><code>SM6_L<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_oxygen_pair_frequency_distance_3<\/code><\/th>\n<td><code>F03[C-O]<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>mean_sanderson_electronegativity_carbon_scaled<\/code><\/th>\n<td><code>Me<\/code><\/td>\n<td>Electronic\/E-state<\/td>\n<\/tr>\n<tr>\n<th><code>mean_first_ionization_potential_carbon_scaled<\/code><\/th>\n<td><code>Mi<\/code><\/td>\n<td>Electronic\/E-state<\/td>\n<\/tr>\n<tr>\n<th><code>hydrazine_nitrogen_count<\/code><\/th>\n<td><code>nN-N<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>aromatic_nitro_group_count<\/code><\/th>\n<td><code>nArNO2<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>crx3_group_count<\/code><\/th>\n<td><code>nCRX3<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>burden_polarizability_normalized_positive_spectral_sum<\/code><\/th>\n<td><code>SpPosA_B(p)<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>circuit_count<\/code><\/th>\n<td><code>nCIR<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_bromine_pair_distance_1_present<\/code><\/th>\n<td><code>B01[C-Br]<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_chlorine_pair_distance_3_present<\/code><\/th>\n<td><code>B03[C-Cl]<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>n073_nitrogen_fragment_count<\/code><\/th>\n<td><code>N-073<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>adjacency_leading_eigenvalue<\/code><\/th>\n<td><code>SpMax_A<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>intrinsic_state_pseudoconnectivity_1d<\/code><\/th>\n<td><code>Psi_i_1d<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_bromine_pair_distance_4_present<\/code><\/th>\n<td><code>B04[C-Br]<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>double_bonded_oxygen_e_state_sum<\/code><\/th>\n<td><code>SdO<\/code><\/td>\n<td>Electronic\/E-state<\/td>\n<\/tr>\n<tr>\n<th><code>laplace_second_mohar_index<\/code><\/th>\n<td><code>TI2_L<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>ring_tertiary_sp3_carbon_count<\/code><\/th>\n<td><code>nCrt<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>r_cx_r_fragment_count<\/code><\/th>\n<td><code>C-026<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>carbon_nitrogen_pair_frequency_distance_2<\/code><\/th>\n<td><code>F02[C-N]<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>hydrogen_bond_donor_atom_count<\/code><\/th>\n<td><code>nHDon<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>burden_mass_leading_eigenvalue<\/code><\/th>\n<td><code>SpMax_B(m)<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>intrinsic_state_pseudoconnectivity_type_s_average<\/code><\/th>\n<td><code>Psi_i_A<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>nitrogen_atom_count<\/code><\/th>\n<td><code>nN<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<tr>\n<th><code>burden_mass_spectral_moment_order_6<\/code><\/th>\n<td><code>SM6_B(m)<\/code><\/td>\n<td>Topological\/spectral<\/td>\n<\/tr>\n<tr>\n<th><code>aromatic_ester_count<\/code><\/th>\n<td><code>nArCOOR<\/code><\/td>\n<td>Functional\/fragment<\/td>\n<\/tr>\n<tr>\n<th><code>halogen_atom_count<\/code><\/th>\n<td><code>nX<\/code><\/td>\n<td>Constitutional<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<\/details>\n<div class=\"nds-figure-grid\">\n<figure><img decoding=\"async\" src=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-class-distribution-2026.png\" alt=\"Class distribution with 356 readily biodegradable and 699 not readily biodegradable records\"><figcaption>The full data set is imbalanced: 33.7% positive and 66.3% negative.<\/figcaption><\/figure>\n<figure><img decoding=\"async\" src=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-input-correlations-2026.png\" alt=\"Univariate descriptor associations with the ready-biodegradability label\"><figcaption>These are one-descriptor associations. They are not causal effects, mechanistic explanations or multivariable feature importance.<\/figcaption><\/figure>\n<\/div>\n<div class=\"nds-note nds-note--provenance\"><strong>Provenance.<\/strong> The data are the UCI <a href=\"https:\/\/archive.ics.uci.edu\/dataset\/254\/qsar+biodegradation\">QSAR biodegradation data set<\/a>, derived from the study by <a href=\"https:\/\/pubs.acs.org\/doi\/10.1021\/ci4000213\">Mansouri et al.<\/a> The UCI table supplies descriptors and the binary endpoint but no CAS identifiers, SMILES, structures, individual test protocols or descriptor-generation workflow. Those omissions constrain auditability and prospective use.<\/div>\n<\/section>\n<section id=\"3-model\" class=\"nds-card\">\n<h2>3. Model<\/h2>\n<p>Thirty-eight descriptor inputs use mean-and-standard-deviation scaling; the three binary bromine\/chlorine indicators use minimum\u2013maximum scaling. The scaled inputs connect directly to one sigmoid output, so this is a nonlinear-preprocessing plus logistic classification model with 42 trainable parameters and no hidden layer.<\/p>\n<p>The output is named <code>readily_biodegradable<\/code>. A larger value supports the positive class, but the sigmoid score has not been calibrated as a probability.<\/p>\n<div class=\"nds-note\"><strong>Output contract.<\/strong> Supply all 41 descriptors in the published order. Do not supply a chemical name, formula or SMILES string directly, and do not report the score as regulatory confidence.<\/div>\n<figure class=\"nds-architecture-figure\" data-model-stage=\"initial\"><img decoding=\"async\" class=\"nds-architecture\" src=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-network-architecture-2026.png\" alt=\"Initial and final direct sigmoid architecture with 41 molecular descriptor inputs and one ready-biodegradability output\"><figcaption>The same direct 41\u20131 architecture is used from initialization through final testing; no hidden-layer or neuron-selection experiment was performed.<\/figcaption><\/figure>\n<\/section>\n<section id=\"4-training\" class=\"nds-card\">\n<h2>4. Training strategy<\/h2>\n<p>The model minimizes a class-weighted squared error with the quasi-Newton method and no explicit regularization. The positive-class weight is 1.4817 and the negative-class weight is 0.7546, compensating partly for the class imbalance.<\/p>\n<p>Training error falls from 1.1168 to 0.2722 over 91 completed epochs. Selection error reaches its minimum of approximately 0.257 at epoch 16 and later rises to 0.3447, while training error continues to improve.<\/p>\n<figure class=\"nds-centered-figure\"><img decoding=\"async\" src=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-training-history-2026.png\" alt=\"Training and selection weighted-squared-error histories over 91 epochs\"><figcaption>The divergence after the selection minimum indicates overfitting. The downloadable export is the stored final model, not an automatically restored epoch-16 checkpoint.<\/figcaption><\/figure>\n<div class=\"nds-note nds-note--warning\"><strong>Training implication.<\/strong> A production rerun should preserve and compare the lowest-selection-error checkpoint, then lock the testing set until the model and threshold are finalized.<\/div>\n<\/section>\n<section id=\"5-selection\" class=\"nds-card\">\n<h2>5. Model selection and baseline<\/h2>\n<p><strong>No neuron selection or architecture selection was performed.<\/strong> The direct 41\u20131 classifier is both the base and final architecture. The selection subset was used to monitor optimization, not to search across model sizes.<\/p>\n<div class=\"nds-table-scroll\">\n<table>\n<thead>\n<tr>\n<th>Reference<\/th>\n<th>Testing accuracy<\/th>\n<th>Balanced accuracy<\/th>\n<th>Interpretation<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th>Always predict the majority class<\/th>\n<td>65.4%<\/td>\n<td>50.0%<\/td>\n<td>Null decision rule for the testing prevalence<\/td>\n<\/tr>\n<tr>\n<th>Final direct sigmoid model<\/th>\n<td>88.2%<\/td>\n<td>86.7%<\/td>\n<td>Material improvement on this internal split<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p>The simple model is a useful transparent baseline. More complexity is not justified until checkpointing, external validation and applicability-domain controls have been addressed.<\/p>\n<\/section>\n<section id=\"6-validation\" class=\"nds-card\">\n<h2>6. Scientific validation<\/h2>\n<p>The held-out testing subset contains 73 positive and 138 negative chemicals. At the default 0.50 threshold, the model classifies 186 of 211 correctly. The ROC analysis evaluates ranking across all thresholds and reports an AUC of 0.931 with a 95% interval from 0.896 to 0.966.<\/p>\n<div class=\"nds-table-scroll\">\n<table class=\"nds-metrics\">\n<thead>\n<tr>\n<th>Testing metric<\/th>\n<th>Value<\/th>\n<th>Operational meaning at threshold 0.50<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th>ROC AUC<\/th>\n<td>0.931<\/td>\n<td>Threshold-independent ranking reported by Neural Designer<\/td>\n<\/tr>\n<tr>\n<th>Accuracy<\/th>\n<td>88.15%<\/td>\n<td>186 correct classifications from 211<\/td>\n<\/tr>\n<tr>\n<th>Sensitivity<\/th>\n<td>82.19%<\/td>\n<td>60 of 73 positive records detected<\/td>\n<\/tr>\n<tr>\n<th>Specificity<\/th>\n<td>91.30%<\/td>\n<td>126 of 138 negative records rejected<\/td>\n<\/tr>\n<tr>\n<th>Precision<\/th>\n<td>83.33%<\/td>\n<td>60 of 72 positive calls correct<\/td>\n<\/tr>\n<tr>\n<th>F1 score<\/th>\n<td>82.76%<\/td>\n<td>Harmonic balance of precision and sensitivity<\/td>\n<\/tr>\n<tr>\n<th>Balanced accuracy<\/th>\n<td>86.75%<\/td>\n<td>Equal weighting of sensitivity and specificity<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<figure class=\"nds-centered-roc\"><img decoding=\"async\" src=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-roc-2026.png\" alt=\"ROC curve for the held-out biodegradation testing subset\"><figcaption>ROC AUC = 0.931. Threshold selection must follow the intended screening cost, not the visual optimum alone.<\/figcaption><\/figure>\n<h3>Confusion matrix at threshold 0.50<\/h3>\n<div class=\"nds-table-scroll\">\n<table class=\"nds-confusion\">\n<thead>\n<tr>\n<th>Actual \/ predicted<\/th>\n<th>Readily biodegradable<\/th>\n<th>Not readily biodegradable<\/th>\n<th>Total<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th>Readily biodegradable<\/th>\n<td>60<\/td>\n<td>13<\/td>\n<td>73<\/td>\n<\/tr>\n<tr>\n<th>Not readily biodegradable<\/th>\n<td>12<\/td>\n<td>126<\/td>\n<td>138<\/td>\n<\/tr>\n<tr>\n<th>Total<\/th>\n<td>72<\/td>\n<td>139<\/td>\n<td>211<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<div class=\"nds-figure-grid\"><\/div>\n<div class=\"nds-note\"><strong>Scientific interpretation.<\/strong> The model separates the two labels well on this random internal split. It has not been tested on an independent chemical collection or on a structure-aware split, so performance on new scaffolds and chemical families remains unknown.<\/div>\n<\/section>\n<section id=\"7-inference\" class=\"nds-card\">\n<h2>7. Inference and reproducibility<\/h2>\n<p>A professional workflow starts from an identified chemical structure, calculates the same descriptor definitions with a versioned toolchain, verifies schema and applicability, produces a score, applies a pre-approved operating threshold and routes the result to expert or laboratory confirmation.<\/p>\n<div class=\"nds-flow\">\n<div>Verified structure<\/div>\n<div>Versioned descriptor calculation<\/div>\n<div>Schema and domain checks<\/div>\n<div>Model score<\/div>\n<div>Operating threshold<\/div>\n<div>Expert or OECD 301 follow-up<\/div>\n<\/div>\n<h3>Screening-threshold scenario<\/h3>\n<p>Changing the threshold does not improve the model; it changes the balance between missed positive chemicals and additional positive calls. A sensitivity-oriented screen can use the ROC-derived 0.29 operating point to miss fewer readily biodegradable records, at the cost of more false positives.<\/p>\n<div class=\"nds-table-scroll\">\n<table>\n<thead>\n<tr>\n<th>Threshold<\/th>\n<th>TP<\/th>\n<th>FN<\/th>\n<th>FP<\/th>\n<th>TN<\/th>\n<th>Sensitivity<\/th>\n<th>Specificity<\/th>\n<th>Precision<\/th>\n<th>Accuracy<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<th>0.50 default<\/th>\n<td>60<\/td>\n<td>13<\/td>\n<td>12<\/td>\n<td>126<\/td>\n<td>82.2%<\/td>\n<td>91.3%<\/td>\n<td>83.3%<\/td>\n<td>88.2%<\/td>\n<\/tr>\n<tr>\n<th>0.29 sensitivity-oriented<\/th>\n<td>64<\/td>\n<td>9<\/td>\n<td>19<\/td>\n<td>119<\/td>\n<td>87.7%<\/td>\n<td>86.2%<\/td>\n<td>77.1%<\/td>\n<td>86.7%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p><strong>Result.<\/strong> The 0.29 threshold detects four additional positives in this testing subset, while creating seven additional false-positive calls. Which trade-off is acceptable depends on whether the next step is inexpensive prioritization or a consequential decision.<\/p>\n<h3>Illustrative exported-model calculation<\/h3>\n<div class=\"nds-deployment-grid\">\n<div>\n<p>The descriptor vector packaged with the export returns a score of <strong>0.11151<\/strong>. It is below both demonstrated thresholds and is therefore assigned to the not-readily-biodegradable class in either scenario.<\/p>\n<p>The vector is not a row of the published table and has no supplied chemical identifier. It demonstrates the software interface only; it is not an extra validation compound.<\/p>\n<\/div>\n<div class=\"nds-result\"><span>Exported sigmoid score<\/span><strong>0.11151<\/strong><\/p>\n<p>Classification: not readily biodegradable at 0.50 and 0.29.<\/p>\n<\/div>\n<\/div>\n<h3>Reproduce the calculation<\/h3>\n<p>The Python package includes the exact model, ordered input schema, illustrative descriptor vector and expected score. The Neural Designer package preserves the split, trained parameters and regenerated analyses.<\/p>\n<pre class=\"nds-code\"><code>from model import NeuralNetwork\n\nscore = NeuralNetwork().calculate_outputs(descriptors)[0]\nscreening_class = int(score &gt;= 0.50)<\/code><\/pre>\n<div class=\"nds-downloads\">\n<a href=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-python-model-2026.zip\">Download Python model (ZIP)<\/a><br \/>\n<a href=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation-neural-designer-project-2026.zip\">Download Neural Designer project (ZIP)<\/a><br \/>\n<a href=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2026\/08\/biodegradation.csv\">Download biodegradation.csv<\/a>\n<\/div>\n<\/section>\n<section id=\"8-validity\" class=\"nds-card\">\n<h2>8. Validity, uncertainty and limitations<\/h2>\n<ul>\n<li><strong>Internal random split only.<\/strong> The 633\/211\/211 partition does not test generalization to unseen scaffolds, chemical families or a later external collection.<\/li>\n<li><strong>No applicability domain.<\/strong> The project does not define leverage, similarity, descriptor-range or structural-alert criteria for accepting a prediction.<\/li>\n<li><strong>Incomplete chemical traceability.<\/strong> The UCI table does not include CAS numbers, SMILES, structures, assay-level metadata or the descriptor-calculation implementation required to reproduce a prediction from a named substance.<\/li>\n<li><strong>Endpoint abstraction.<\/strong> The binary target aggregates source ready-biodegradability outcomes; it is not a direct simulation of one fully documented OECD 301 protocol.<\/li>\n<li><strong>Uncalibrated scores.<\/strong> The sigmoid output is useful for ranking and thresholding but is not demonstrated to be a calibrated probability.<\/li>\n<li><strong>Training\/checkpoint risk.<\/strong> Selection error is lowest near epoch 16, whereas the stored export comes from the later final epoch.<\/li>\n<li><strong>Regulatory boundary.<\/strong> OECD guidance expects a defined endpoint and algorithm, an applicability domain, appropriate predictivity measures and, where possible, mechanistic interpretation. This tutorial does not satisfy that complete package.<\/li>\n<\/ul>\n<div class=\"nds-note nds-note--warning\"><strong>Decision boundary.<\/strong> Use the model for reproducible research and pre-screening. Confirm consequential conclusions with documented expert review and appropriate experimental evidence.<\/div>\n<\/section>\n<section id=\"references\" class=\"nds-card\">\n<h2>References<\/h2>\n<ul>\n<li><a href=\"https:\/\/archive.ics.uci.edu\/dataset\/254\/qsar+biodegradation\">UCI Machine Learning Repository: QSAR biodegradation<\/a>.<\/li>\n<li>Mansouri K, Ringsted T, Ballabio D, Todeschini R, Consonni V. <a href=\"https:\/\/pubs.acs.org\/doi\/10.1021\/ci4000213\">Quantitative Structure\u2013Activity Relationship Models for Ready Biodegradability of Chemicals<\/a>. <em>Journal of Chemical Information and Modeling<\/em>. 2013;53(4):867\u2013878.<\/li>\n<li><a href=\"https:\/\/www.oecd.org\/en\/publications\/test-no-301-ready-biodegradability_9789264070349-en.html\">OECD Test No. 301: Ready Biodegradability<\/a>.<\/li>\n<li><a href=\"https:\/\/www.oecd.org\/en\/publications\/guidance-document-on-the-validation-of-quantitative-structure-activity-relationship-q-sar-models_9789264085442-en.html\">OECD Guidance Document on the Validation of (Q)SAR Models<\/a>.<\/li>\n<li>Sahigara F et al. <a href=\"https:\/\/pmc.ncbi.nlm.nih.gov\/articles\/PMC5382130\/\">Comparison of different approaches to define the applicability domain of QSAR models<\/a>. <em>Molecules<\/em>. 2012;17:4791\u20134810.<\/li>\n<\/ul>\n<\/section>\n<\/div>\n<\/div>\n","protected":false},"author":13,"featured_media":2555,"template":"","categories":[29],"tags":[40,43],"class_list":["post-3519","learning","type-learning","status-publish","has-post-thumbnail","hentry","category-examples","tag-chemistry","tag-industry"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v26.4 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Predict chemical biodegradability with machine learning<\/title>\n<meta name=\"description\" content=\"Build a machine learning model that predicts the biodegradability of molecules by analyzing their chemical structures and properties.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"QSAR Biodegradation machine learning example\" \/>\n<meta property=\"og:description\" content=\"This example aims to catalog different chemicals depending on whether they are biodegradable or not. QSAR (Quantitative Structure Activity Relationships) models are currently being developed with more frequency nowadays.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/\" \/>\n<meta property=\"og:site_name\" content=\"Neural Designer\" \/>\n<meta property=\"article:modified_time\" content=\"2026-08-07T09:44:05+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"628\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:title\" content=\"QSAR Biodegradation machine learning example\" \/>\n<meta name=\"twitter:description\" content=\"This example aims to catalog different chemicals depending on whether they are biodegradable or not. QSAR (Quantitative Structure Activity Relationships) models are currently being developed with more frequency nowadays.\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp\" \/>\n<meta name=\"twitter:site\" content=\"@NeuralDesigner\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"7 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/\",\"url\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/\",\"name\":\"Predict chemical biodegradability with machine learning\",\"isPartOf\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp\",\"datePublished\":\"2023-08-31T11:12:58+00:00\",\"dateModified\":\"2026-08-07T09:44:05+00:00\",\"description\":\"Build a machine learning model that predicts the biodegradability of molecules by analyzing their chemical structures and properties.\",\"breadcrumb\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage\",\"url\":\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp\",\"contentUrl\":\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp\",\"width\":1200,\"height\":628,\"caption\":\"Biodegradation\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.neuraldesigner.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Learning\",\"item\":\"https:\/\/www.neuraldesigner.com\/learning\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Predict chemical biodegradability with machine learning\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.neuraldesigner.com\/#website\",\"url\":\"https:\/\/www.neuraldesigner.com\/\",\"name\":\"Neural Designer\",\"description\":\"Explanable AI Platform\",\"publisher\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.neuraldesigner.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.neuraldesigner.com\/#organization\",\"name\":\"Neural Designer\",\"url\":\"https:\/\/www.neuraldesigner.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.neuraldesigner.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/05\/logo-neural-1.png\",\"contentUrl\":\"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/05\/logo-neural-1.png\",\"width\":1024,\"height\":223,\"caption\":\"Neural Designer\"},\"image\":{\"@id\":\"https:\/\/www.neuraldesigner.com\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/x.com\/NeuralDesigner\",\"https:\/\/es.linkedin.com\/showcase\/neuraldesigner\/\"]}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Predict chemical biodegradability with machine learning","description":"Build a machine learning model that predicts the biodegradability of molecules by analyzing their chemical structures and properties.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/","og_locale":"en_US","og_type":"article","og_title":"QSAR Biodegradation machine learning example","og_description":"This example aims to catalog different chemicals depending on whether they are biodegradable or not. QSAR (Quantitative Structure Activity Relationships) models are currently being developed with more frequency nowadays.","og_url":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/","og_site_name":"Neural Designer","article_modified_time":"2026-08-07T09:44:05+00:00","og_image":[{"width":1200,"height":628,"url":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp","type":"image\/webp"}],"twitter_card":"summary_large_image","twitter_title":"QSAR Biodegradation machine learning example","twitter_description":"This example aims to catalog different chemicals depending on whether they are biodegradable or not. QSAR (Quantitative Structure Activity Relationships) models are currently being developed with more frequency nowadays.","twitter_image":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp","twitter_site":"@NeuralDesigner","twitter_misc":{"Est. reading time":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/","url":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/","name":"Predict chemical biodegradability with machine learning","isPartOf":{"@id":"https:\/\/www.neuraldesigner.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage"},"image":{"@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage"},"thumbnailUrl":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp","datePublished":"2023-08-31T11:12:58+00:00","dateModified":"2026-08-07T09:44:05+00:00","description":"Build a machine learning model that predicts the biodegradability of molecules by analyzing their chemical structures and properties.","breadcrumb":{"@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#primaryimage","url":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp","contentUrl":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/06\/biodegradation-example.webp","width":1200,"height":628,"caption":"Biodegradation"},{"@type":"BreadcrumbList","@id":"https:\/\/www.neuraldesigner.com\/learning\/examples\/qsar-biodegradation\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.neuraldesigner.com\/"},{"@type":"ListItem","position":2,"name":"Learning","item":"https:\/\/www.neuraldesigner.com\/learning\/"},{"@type":"ListItem","position":3,"name":"Predict chemical biodegradability with machine learning"}]},{"@type":"WebSite","@id":"https:\/\/www.neuraldesigner.com\/#website","url":"https:\/\/www.neuraldesigner.com\/","name":"Neural Designer","description":"Explanable AI Platform","publisher":{"@id":"https:\/\/www.neuraldesigner.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.neuraldesigner.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.neuraldesigner.com\/#organization","name":"Neural Designer","url":"https:\/\/www.neuraldesigner.com\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.neuraldesigner.com\/#\/schema\/logo\/image\/","url":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/05\/logo-neural-1.png","contentUrl":"https:\/\/www.neuraldesigner.com\/wp-content\/uploads\/2023\/05\/logo-neural-1.png","width":1024,"height":223,"caption":"Neural Designer"},"image":{"@id":"https:\/\/www.neuraldesigner.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/NeuralDesigner","https:\/\/es.linkedin.com\/showcase\/neuraldesigner\/"]}]}},"_links":{"self":[{"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/learning\/3519","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/learning"}],"about":[{"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/types\/learning"}],"author":[{"embeddable":true,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/users\/13"}],"version-history":[{"count":5,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/learning\/3519\/revisions"}],"predecessor-version":[{"id":23156,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/learning\/3519\/revisions\/23156"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/media\/2555"}],"wp:attachment":[{"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/media?parent=3519"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/categories?post=3519"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.neuraldesigner.com\/api\/wp\/v2\/tags?post=3519"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}