Abstract:e13597 Background: Accurate identification of tumor origin is crucial for effective diagnosis and treatment, particularly in cases of metastatic tumors. Interpretable machine learning models have displayed significant potential in addressing this problem when imaging and immunohistochemistry (IHC) examinations are ineffective. In this study, we developed a panel using gene expression profiles from various tumors and constructed a robust machine learning model for precise identification of tumor origin. Methods: RNA sequencing (RNA-seq) data of 9462 tumor samples originating from 21 different organs were collected from The Cancer Genome Atlas (TCGA). We conducted feature engineering through unsupervised clustering and differential gene expression analysis, selecting a refined panel of 164 genes from a pool of over 60,000 identifiers. Subsequently, a machine learning classifier grounded in Logistic regression (LR) was trained on 9462 samples with the constructed 164-gene panel. To enhance the adaptability of our model to this task, 10-fold cross-validation was employed in the multi-class mode. Two independent test sets, the Primary tumor set (PT, n=3420, including 19 tumor types) and the Metastatic tumor set (MTP, n=100, all originating from the prostate and spread to bone, liver, etc.), were established with samples from Gene Expression Omnibus (GEO) and other published studies. Notably, all samples underwent FPKM normalization and there was no overlap between training and testing samples. Model performance was assessed using accuracy, specificity, and sensitivity metrics. Results: The 164-gene expression panel achieved a cross-validation accuracy of 96.73%. In the assessment of the PT test set, our model achieved an overall 99.62% specificity, 92.31% accuracy and 89.79% sensitivity, exhibiting performance comparable to similar models in other studies. Additionally, our model accurately traced 91% of metastatic tumors in the MTP test set to the prostate, surpassing previous lines of work by a large margin. Conclusions: Gene expression patterns reveal organ-specific characteristics that could be used to identify tumor origin. The combination of a condensed yet comprehensive gene expression panel with a robust machine learning model serves as a promising tool for tumor diagnosis. Ongoing correlative studies aim to extend predictions from tissue organs to cancer subtypes.

Using 2d principal component analysis to reduce dimensionality of gene expression profiles for tumor classification

RPCA-Based Tumor Classification Using Gene Expression Data

A Dynamic Subspace Learning Method for Tumor Classification Using Microarray Gene Expression Data

Dual Hypergraph Regularized PCA for Biclustering of Tumor Gene Expression Data

Enhancing Characteristic Gene Selection and Tumor Classification by the Robust Laplacian Supervised Discriminative Sparse PCA

Supervised Discriminative Sparse PCA for Com-Characteristic Gene Selection and Tumor Classification on Multiview Biological Data

Comparison among dimensionality reduction techniques based on Random Projection for cancer classification.

PCA AND PLS FOR GASTRIC CANCER SUBTYPE CLASSIFICATION

Multi-Class Cancer Classification by Total Principal Component Regression (tpcr) Using Microarray Gene Expression Data

Feature (gene) Selection in Gene Expression-Based Tumor Classification

Nonlinear dimensionality reduction of gene expression data for visualization and clustering analysis of cancer tissue samples.

Deep Learning Based Tumor Type Classification Using Gene Expression Data

Based on an Improved Pre-Pca + LDA Classifier Design in Tumor Cells

Cancer classification from the gene expression profiles by Discriminant Kernel-PLS.

Cancer Classification Using Entropy Analysis in Fractional Fourier Domain of Gene Expression Profile

Discriminant Projection Shared Dictionary Learning for Classification of Tumors Using Gene Expression Data.

PCA via joint graph Laplacian and sparse constraint: Identification of differentially expressed genes and sample clustering on gene expression data

Tumor origin identification through machine learning and gene expression profiling.

Sparse Representation for Tumor Classification Based on Feature Extraction Using Latent Low-Rank Representation.

ESTIMATING THE NUMBER OF CANCER SUBTYPES FROM WHOLE-GENOME EXPRESSION DATA VIA A PENALIZED PROBABILISTIC PRINCIPAL COMPONENT ANALYSIS ∗ By

Informative gene selection and tumor classification by null space LDA for microarray data