Marleen van Greevenbroek
26 november 2020 bij 18:36- Antwoorden
Thank you for this presentation. Interesting setup ! Minor question of the P value: how was the Bonferoni corrction derived (0.05/99=0.0005)
W. Wang (Leiden)
26 november 2020 bij 19:29- Antwoorden
Hi Marleen, thank you for your question!
Actually, the Bonferoni correction derived from (0.05/37=0.00135). Since conventional Bonferroni correction is too stringent given the high correlation between multiple of the included metabolomic measures, we corrected for the number of independent metabolomic measures by using the methodology that published by Li et al (Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix.), which derived the number 37.
2 Reacties
Thank you for this presentation. Interesting setup ! Minor question of the P value: how was the Bonferoni corrction derived (0.05/99=0.0005)
Hi Marleen, thank you for your question!
Actually, the Bonferoni correction derived from (0.05/37=0.00135). Since conventional Bonferroni correction is too stringent given the high correlation between multiple of the included metabolomic measures, we corrected for the number of independent metabolomic measures by using the methodology that published by Li et al (Adjusting multiple testing in multilocus analyses using the eigenvalues of a correlation matrix.), which derived the number 37.