Statistical Issues and Analyses of in vivo and in vitro Genomic Data in order to Identify Clinically Relevant Profiles
2007

Integrating In Vitro and In Vivo Genomic Data in Cancer Research

Sample size: 295 publication Evidence: moderate

Author Information

Author(s): Laila M. Poisson, Debashis Ghosh

Primary Institution: University of Michigan

Hypothesis

Can gene signatures from in vitro experiments predict clinical outcomes in in vivo cancer samples?

Conclusion

The study demonstrates that in vitro gene signatures can effectively predict clinical outcomes in various cancer types.

Supporting Evidence

  • The in vitro gene signature was derived from a wound healing study.
  • Significant correlations were found between in vitro and in vivo gene expression profiles.
  • Permutation testing showed that the in vitro signature outperformed random gene sets.

Takeaway

Scientists are trying to see if results from lab tests on cells can help predict how real tumors behave in patients.

Methodology

The study used gene expression profiling and statistical tests to compare in vitro and in vivo data.

Potential Biases

Potential bias due to the selection of genes and the use of different datasets.

Limitations

The study's findings may be limited by the differences in microarray platforms and gene mapping.

Participant Demographics

The study involved various cancer types including prostate, breast, and lung cancers.

Statistical Information

P-Value

p < 0.0001 for prostate tumors, p = 0.0207 for breast tumors, p = 0.0352 for lung tumors

Statistical Significance

p<0.05

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication