Automatic discovery of cross-family sequence features associated with protein function
2006

Discovering Protein Function from Sequences

Sample size: 2146 publication Evidence: moderate

Author Information

Author(s): Markus Brameier, Josien Haan, Andrea Krings, Robert M. MacCallum

Primary Institution: Stockholm Bioinformatics Center, Stockholm University

Hypothesis

Can a self-supervised data mining approach identify relationships between protein sequences and their functions without predefined categories?

Conclusion

The study developed a novel approach that successfully identifies functionally important sequence features without requiring expert knowledge.

Supporting Evidence

  • The strongest correlations were found between sequence features and annotations related to cellular compartments.
  • The study identified novel sequence-to-function relationships, such as the role of polyglutamine in transcription.
  • The self-supervised approach outperformed control experiments, indicating its effectiveness in discovering functional relationships.

Takeaway

This study shows a new way to find out what proteins do just by looking at their sequences, without needing experts to tell us first.

Methodology

The study used a self-supervised data mining approach with genetic programming to analyze protein sequences and their annotations.

Potential Biases

The dataset may have low contrast between positive and negative examples, affecting prediction accuracy.

Limitations

The predictors had low specificity and sensitivity, making them less useful for biologists.

Statistical Information

P-Value

p<0.001

Statistical Significance

p<0.001

Digital Object Identifier (DOI)

10.1186/1471-2105-7-16

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication