Support Vector Machines and Kernels for Computational Biology
2008

Support Vector Machines and Kernels for Computational Biology

Sample size: 2200 publication Evidence: moderate

Author Information

Author(s): Ben-Hur Asa, Ong Cheng Soon, Sonnenburg Sören, Schölkopf Bernhard, Rätsch Gunnar

Primary Institution: Colorado State University

Hypothesis

Can support vector machines and kernel methods effectively solve prediction problems in computational biology?

Conclusion

Support vector machines (SVMs) and kernel methods are effective for solving various prediction problems in computational biology, particularly in recognizing splice sites.

Supporting Evidence

  • SVMs are widely used in computational biology due to their high accuracy.
  • The study demonstrated the effectiveness of SVMs in recognizing splice sites.
  • Different kernel functions can improve classifier performance.

Takeaway

This study shows how computers can help scientists find important parts of genes by using special math tools called support vector machines.

Methodology

The study used support vector machines with various kernel functions to classify splice sites based on features derived from biological sequences.

Limitations

The dataset used for training was smaller and less unbalanced than typical datasets in the field.

Digital Object Identifier (DOI)

10.1371/journal.pcbi.1000173

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication