{"data":[{"id":"10.5061/dryad.z08kprrx0","type":"dois","attributes":{"doi":"10.5061/dryad.z08kprrx0","identifiers":[],"creators":[{"nameType":"Personal","affiliation":["Philippine Nuclear Research Institute","De La Salle University"],"name":"Hila, Abigaile Mia","nameIdentifiers":[{"nameIdentifierScheme":"ORCID","schemeUri":"https://orcid.org","nameIdentifier":"https://orcid.org/0000-0001-7127-586X"}]},{"nameType":"Personal","affiliation":["De La Salle University"],"name":"Amalin, Divina","nameIdentifiers":[]},{"nameType":"Personal","affiliation":["Philippine Nuclear Research Institute"],"name":"Obra, Glenda","nameIdentifiers":[]},{"nameType":"Personal","affiliation":["De La Salle University"],"name":"Almarinez, Billy Joel","nameIdentifiers":[]},{"nameType":"Personal","affiliation":["De La Salle University","St. Luke's Medical Center"],"name":"Enriquez, Ma. Luisa","nameIdentifiers":[]},{"nameType":"Personal","affiliation":["De La Salle University"],"name":"Carvajal, Thaddeus","nameIdentifiers":[]}],"titles":[{"title":"Integrating developmental traits, Morphometrics, and machine learning for sex classification of \u003cem\u003eAedes aegypti\u003c/em\u003e (Diptera: Culicidae)"}],"publisher":"Dryad","container":{},"publicationYear":2026,"subjects":[{"schemeUri":"https://web-archive.oecd.org/2012-06-15/138575-38235147.pdf","subject":"FOS: Biological sciences","subjectScheme":"fos"},{"schemeUri":"https://github.com/PLOS/plos-thesaurus","subject":"Artificial intelligence","subjectScheme":"PLOS Subject Area Thesaurus"},{"schemeUri":"https://github.com/PLOS/plos-thesaurus","subject":"Deep learning","subjectScheme":"PLOS Subject Area Thesaurus"},{"subject":"Dengue vector"},{"subject":"Gaussian curves"},{"subject":"Mosquito pupal sex sorting"},{"schemeUri":"https://github.com/PLOS/plos-thesaurus","subject":"Sexual dimorphism","subjectScheme":"PLOS Subject Area Thesaurus"}],"contributors":[{"nameType":"Personal","affiliation":["Philippine Nuclear Research Institute","De La Salle University"],"name":"Hila, Abigaile Mia","nameIdentifiers":[{"nameIdentifierScheme":"ORCID","schemeUri":"https://orcid.org","nameIdentifier":"https://orcid.org/0000-0001-7127-586X"}],"contributorType":"ContactPerson"},{"nameType":"Personal","affiliation":["De La Salle University"],"name":"Carvajal, Thaddeus","contributorType":"ContactPerson","nameIdentifiers":[]}],"dates":[{"date":"2026-08-22T16:18:40Z","dateType":"Created"},{"date":"2026-10-09T09:17:36Z","dateType":"Submitted"},{"date":"2026-10-14T00:00:00Z","dateType":"Issued"},{"date":"2026-10-14T00:00:00Z","dateType":"Available"}],"language":"en","types":{"schemaOrg":"Dataset","resourceTypeGeneral":"Dataset","citeproc":"dataset","bibtex":"misc","ris":"DATA","resourceType":"dataset"},"relatedIdentifiers":[{"relationType":"IsDerivedFrom","relatedIdentifier":"10.5281/zenodo.16986111","relatedIdentifierType":"DOI"}],"relatedItems":[],"sizes":["118680 bytes"],"formats":[],"version":"3","rightsList":[{"rightsIdentifierScheme":"SPDX","rightsUri":"https://creativecommons.org/publicdomain/zero/1.0/legalcode","schemeUri":"https://spdx.org/licenses/","rights":"Creative Commons Zero v1.0 Universal","rightsIdentifier":"cc0-1.0"}],"descriptions":[{"descriptionType":"Abstract","description":"This dataset contains raw morphometric measurements, developmental\n tracking records, and supervised machine learning classification\n performance reports for the dengue vector Aedes aegypti (Old Balara\n strain) to support sex differentiation research in Sterile Insect\n Technique (SIT) and mass-rearing workflows. Across six consecutive\n laboratory generations (F97 – F102), developmental traits and eight\n morphometric parameters (cephalothorax length, width, perimeter, area,\n pupal length, pupal weight, adult wing length, and thorax length) were\n systematically recorded from 600 individual specimens (N = 600; 300 males,\n 300 females). The repository includes the primary experimental data\n (Measurements.csv) containing specimen-level phenotypic, developmental,\n and verified adult sex records, alongside comprehensive classification\n performance reports (.xlsx format). These reports document precision,\n recall, F1-score, accuracy, and support across nine supervised machine\n learning algorithms evaluated under an 80:20 stratified train-test split\n (n = 480 training, n = 120 testing) with 5-fold cross-validation for both\n single pupal traits and multi-feature models. These data provide\n standardized baseline metrics for automated optical sex-sorting\n prototyping, dimorphic feature selection, and comparative entomological\n morphometrics, with all records collected under standard laboratory\n protocols without proprietary or ethical restrictions."},{"descriptionType":"TechnicalInfo","description":"# Integrating developmental traits, Morphometrics, and machine learning\n for sex classification of *Aedes aegypti* (Diptera: Culicidae) Dataset\n DOI: [10.5061/dryad.z08kprrx0](https://doi.org/10.5061/dryad.z08kprrx0) ##\n Description of the data and file structure This dataset was generated to\n investigate sex differentiation in the dengue vector *Aedes aegypti* (Old\n Balara laboratory strain) to support sex-sorting applications in Sterile\n Insect Technique (SIT) and Incompatible Insect Technique (IIT)\n mass-rearing programs. Across six consecutive generations F97 to F102, 600\n individual mosquitoes (50 males and 50 females per generation) were reared\n under standardized laboratory conditions. Morphometric traits across the\n pupal and adult stages—specifically cephalothorax length, width,\n perimeter, area, pupal length, pupal weight, wing length, and thorax\n length, alongside developmental parameters were measured to quantify\n sexual dimorphism. The collected pupal morphometric measurements were\n subsequently used as predictive features to train, cross-validate, and\n test multiple supervised machine learning algorithms (evaluating\n single-feature vs. multi-feature models) to determine the most effective\n automated sex-classification parameters. ### Files and variables ####\n File: Cephalothorax_Area_(mm)_Reports.xlsx **Description:** Classification\n reports for nine machine learning models using single-feature predictor\n **cephalothorax area** on the test cohort (n = 120). ##### Variables *\n Rows (`Female`, `Male`, `accuracy`, `macro avg`, `weighted avg`) and\n Columns (`precision`, `recall`, `f1-score`, `support`), identical in\n structure and definitions to `Overall_Reports_6Pupa_Parameters.xlsx`. ####\n File: Cephalothorax_Length_(mm)_Reports.xlsx\n **Description:** Classification reports for nine machine learning models\n using single-feature predictor **cephalothorax\n length **on the test cohort (n = 120). ##### Variables *\n Rows (`Female`, `Male`, `accuracy`, `macro avg`, `weighted avg`) and\n Columns (`precision`, `recall`, `f1-score`, `support`), identical in\n structure and definitions to `Overall_Reports_6Pupa_Parameters.xlsx`. ####\n File: Cephalothorax_Perimeter_(mm)_Reports.xlsx\n **Description:** Classification reports for nine machine learning models\n using single-feature predictor **cephalothorax perimeter** on the test\n cohort (n = 120). ##### Variables * Rows (`Female`, `Male`, `accuracy`,\n `macro avg`, `weighted avg`) and Columns (`precision`, `recall`,\n `f1-score`, `support`), identical in structure and definitions to\n `Overall_Reports_6Pupa_Parameters.xlsx`. #### File:\n Cephalothorax_Width_(mm)_Reports.xlsx **Description:** Classification\n reports for nine machine learning models using single-feature predictor\n **cephalothorax width** on the test cohort (n = 120). ##### Variables *\n Rows (`Female`, `Male`, `accuracy`, `macro avg`, `weighted avg`) and\n Columns (`precision`, `recall`, `f1-score`, `support`), identical in\n structure and definitions to `Overall_Reports_6Pupa_Parameters.xlsx`. ####\n File: Overall_Reports_6Pupa_Parameters.xlsx\n **Description:** Classification reports (Scikit-Learn format) for nine\n machine learning models (Random Forest, KNN, Deep Learning/MLP, Gradient\n Boosting, Decision Tree, AdaBoost, SVM, Logistic Regression, Gaussian\n Naive Bayes) using all six pupal parameters simultaneously to classify sex\n on the test cohort (n = 120). ##### **Variables (Rows \u0026amp; Columns):** *\n **Class / Metric Row Labels:** * **Female:** Performance metrics for\n female class classification. * **Male:** Performance metrics for male\n class classification. * **accuracy:** Overall classification accuracy\n across all test samples. * **macro avg:** Unweighted mean of metrics\n across both sex classes. * **weighted avg:** Average of metrics weighted\n by the support of each class. * **Columns:** * **precision:** Ratio of\n true positive predictions to total predicted positives for the class\n (range: 0.00–1.00). * **recall:** Sensitivity or true positive rate (ratio\n of true positives to total actual class samples; range: 0.00–1.00). *\n **f1-score:** Harmonic mean of precision and recall ((2× ((Precision x\n Recall) / (Precision + Recall)); range: 0.00–1.00). * **support:** Number\n of actual occurrences/specimens evaluated in the test dataset (n = 60\n females, n = 60 males; total n = 120). #### File: Measurements.csv\n **Description:** Raw morphometric and developmental dataset containing\n individual records for 600 *Aedes aegypti* specimens (300 males and 300\n females) across six consecutive laboratory generations (F97 to F102). \n ##### Variables * **Gen:** Rearing generation of the Old Balara laboratory\n strain (`F97` to `F102`). * **Replication:** Individual specimen\n identifier within each generation batch (numeric integer: 1–100). *\n **Sex:** Verified biological sex of the adult mosquito post-eclosion\n (`Male` or `Female`). * **Day of pupation:** Time elapsed from egg\n hatching to pupal formation (in whole days). * **Cephalothorax length\n (mm):** Maximum linear anterior–posterior length of the pupal\n cephalothorax (measured in millimeters, mm). * **Cephalothorax width\n (mm):** Maximum transverse lateral width across the pupal cephalothorax\n (measured in millimeters, mm). * **Cephalothorax circumference (mm):**\n Total boundary perimeter traced around the 2D dorsal projection of the\n pupal cephalothorax (measured in millimeters, mm). * **Cephalothorax area\n (mm):** Projected 2D surface area of the pupal cephalothorax (measured in\n square millimeters, $\\text{mm}^2$). *(Column header appears as*\n *`Cephalothorax area (mm)`* *in the raw file).* * **Pupal length (mm):**\n Total longitudinal length extending from the anterior crest of the\n cephalothorax to the distal tip of the pupal paddles (measured in\n millimeters, mm). * **Pupal weight (mg):** Individual pupal wet weight\n measured using an analytical microbalance after surface water was blotted\n using a microfiber cloth (measured in milligrams, mg). * **Wing length\n (mm):** Maximum linear length of the adult right wing, measured from the\n axial notch to the apical margin (measured in millimeters, mm). * **Thorax\n length (mm):** Linear dorsal length of the adult thorax measured from the\n anterior scutum to the posterior margin of the scutellum (measured in\n millimeters, mm). #### File: Pupal_Length_(mm)_Reports.xlsx\n **Description:** Classification reports for nine machine learning models\n using single-feature predictor **pupal length **on the\n test cohort (n = 120). ##### Variables * Rows (`Female`, `Male`,\n `accuracy`, `macro avg`, `weighted avg`) and Columns (`precision`,\n `recall`, `f1-score`, `support`), identical in structure and definitions\n to `Overall_Reports_6Pupa_Parameters.xlsx`. #### File:\n Pupal_Weight_(mg)_Reports.xlsx **Description:** Classification reports for\n nine machine learning models using single-feature predictor **pupal\n weight** on the test cohort (n = 120). ##### Variables * Rows (`Female`,\n `Male`, `accuracy`, `macro avg`, `weighted avg`) and Columns (`precision`,\n `recall`, `f1-score`, `support`), identical in structure and definitions\n to `Overall_Reports_6Pupa_Parameters.xlsx`. ## Code/software ###\n **Software Required to View and Process the Data** The data files can be\n accessed, viewed, and analyzed using standard open-source or freely\n available spreadsheet viewers and programming environments: *\n **Spreadsheet and CSV Data Viewing:** * **LibreOffice Calc** (v7.0 or\n higher) or any standard CSV/XLSX viewer (e.g., Microsoft Excel, Google\n Sheets) can be used to view `Measurements.csv` and the machine learning\n report files (`.xlsx`). \n \n ### **Computational and Machine Learning Environment** All machine\n learning modeling, cross-validation, feature importance evaluation, and\n metric generation were implemented in Python. * **Programming Language\n \u0026amp; Core Environment:** * **Python** (version 3.9 or higher) * **Key\n Packages \u0026amp; Dependencies:** * **NumPy** (v1.22+) – Numerical computing\n and array handling * **Pandas** (v1.4+) – Tabular data manipulation and\n CSV processing * **SciPy** (v1.8+) – Statistical computing and descriptive\n statistics * **Scikit-Learn** (`scikit-learn` v1.0+) – Data partitioning,\n machine learning classification algorithms, k-fold cross-validation, and\n classification report generation (`classification_report`,\n `confusion_matrix`) * **Matplotlib** / **Seaborn** – Visualization and\n distribution plotting ### **Data Processing \u0026amp; Analytical Workflow** 1.\n **Data Ingestion \u0026amp; Cleaning:** Raw morphometric measurements and\n developmental records (`Measurements.csv`) were loaded into Python using\n `pandas`. 2. **Dataset Partitioning:** The pooled laboratory dataset ($N =\n 600$; 300 males, 300 females) was partitioned into a **training cohort\n (80%, n = 480)** and an **independent test cohort (20%, n = 120)** using a\n stratified random split (`train_test_split(..., test_size=0.20,\n stratify=y, random_state=42)`). 3. **Model Training \u0026amp;\n Cross-Validation:** Nine supervised algorithms (Random Forest, K-Nearest\n Neighbors, Deep Learning / MLP, Gradient Boosting, Decision Tree,\n AdaBoost, Support Vector Machine, Logistic Regression, and Gaussian Naive\n Bayes) were trained and evaluated on the training set using 5-fold\n stratified cross-validation. 4. **Single-Feature vs. Multi-Feature\n Evaluation:** * Each of the six pupal morphometric features was tested\n individually as a single predictor across all nine algorithms to generate\n individual feature performance reports\n (`Cephalothorax_Area_(mm)_Reports.xlsx`, etc.). * All six pupal features\n were combined into a multi-variable feature matrix to generate\n comprehensive classification metrics\n (`Overall_Reports_6Pupa_Parameters.xlsx`). 5. **Testing \u0026amp; Performance\n Metrics:** Final evaluation metrics (precision, recall, F1-score, overall\n accuracy, macro average, weighted average, and support) were computed\n against the holdout 20% test cohort (n = 120). ## Access information Other\n publicly accessible locations of the data: * The dataset is also archived\n via Zenodo\n Repository: [https://doi.org/10.5281/zenodo.16986111](https://doi.org/10.5281/zenodo.16986111)"}],"geoLocations":[],"fundingReferences":[{"funderIdentifierType":"ROR","funderName":"Science Education Institute","funderIdentifier":"https://ror.org/0240hay17"}],"url":"https://datadryad.org/dataset/doi:10.5061/dryad.z08kprrx0","contentUrl":null,"metadataVersion":0,"schemaVersion":"http://datacite.org/schema/kernel-4","source":"mds","isActive":true,"state":"findable","reason":null,"viewCount":0,"downloadCount":0,"referenceCount":0,"citationCount":0,"partCount":0,"partOfCount":0,"versionCount":0,"versionOfCount":0,"created":"2026-10-14T16:11:56Z","registered":"2026-10-14T16:11:57Z","published":null,"updated":"2026-10-14T16:11:57Z"},"relationships":{"client":{"data":{"id":"dryad.dryad","type":"clients"}}}}],"meta":{"total":170657,"totalPages":10000,"page":1},"links":{"self":"https://api.datacite.org/dois?client_id=dryad.dryad\u0026page%5Bsize%5D=1\u0026affiliation=True\u0026publisher=True","next":"https://api.datacite.org/dois?affiliation=True\u0026client-id=dryad.dryad\u0026page%5Bnumber%5D=2\u0026page%5Bsize%5D=1\u0026publisher=True"}}