17533266. TOPOLOGICAL SIGNATURES FOR DISEASE CHARACTERIZATION simplified abstract (International Business Machines Corporation)

From WikiPatents
Jump to navigation Jump to search

TOPOLOGICAL SIGNATURES FOR DISEASE CHARACTERIZATION

Organization Name

International Business Machines Corporation

Inventor(s)

Laxmi Parida of Mohegan Lake NY (US)

Aldo Guzman Saenz of White Plains NY (US)

Sayan Mandal of Santa Fe NM (US)

Niina Haiminen of Valhalla NY (US)

TOPOLOGICAL SIGNATURES FOR DISEASE CHARACTERIZATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 17533266 titled 'TOPOLOGICAL SIGNATURES FOR DISEASE CHARACTERIZATION

Simplified Explanation

Abstract: This patent application describes a method for predicting the phenotype of a new sample based on gene expression data. The method involves determining pair-wise similarities between genes in the gene expression data and transforming the data into topological summaries. A neural network is then trained using a training set created from these summaries. The trained neural network can predict the phenotype of new samples.

  • Gene expression data associated with a subject is received.
  • Pair-wise similarities between genes in the data are determined.
  • The data is transformed into topological summaries based on the pair-wise similarities.
  • A neural network is trained using a training set created from the topological summaries.
  • A new sample is received and input to the neural network.
  • The neural network predicts the phenotype of the new sample.

Potential Applications:

  • Medical diagnosis and personalized medicine: Predicting the phenotype of a new sample can help in diagnosing diseases and determining the most effective treatment options for individual patients.
  • Drug discovery and development: The method can be used to predict the effects of new drugs on different phenotypes, aiding in the development of targeted therapies.
  • Agricultural research: Predicting the phenotype of plants or animals can assist in breeding programs and improving crop yield or livestock quality.

Problems Solved:

  • Traditional methods of predicting phenotypes based on gene expression data may be time-consuming and less accurate.
  • The method described in the patent application provides a more efficient and accurate way of predicting phenotypes using topological summaries and a trained neural network.

Benefits:

  • Faster and more accurate predictions of phenotypes based on gene expression data.
  • Potential for personalized medicine and targeted therapies.
  • Improved efficiency in drug discovery and agricultural research.


Original Abstract Submitted

Gene expression data associated with a subject can be received. Pair-wise similarities between genes in the gene expression data can be determined. The gene expression data can be transformed into topological summaries based on the pair-wise similarities. A neural network can be trained using a training set created based on the topological summaries. A new sample can be received and input to the neural network, where the neural network can predict the new sample's phenotype.

Retrieved from "http://wikipatents.org/index.php?title=17533266._TOPOLOGICAL_SIGNATURES_FOR_DISEASE_CHARACTERIZATION_simplified_abstract_(International_Business_Machines_Corporation)&oldid=25141"