Minimum sample sizes for identifying chromosomal fragile sites from individuals: Monte Carlo estimation

Area of Science:

• Genetics
• Biostatistics

Background:

• Fragile sites are chromosomal regions prone to breakage.
• Accurate identification of fragile sites is crucial for understanding genomic instability.
• The FSM methodology uses statistical tests to identify fragile sites from breakage data.

Purpose of the Study:

• To estimate the exact level of the standardized X2 test statistic (Xs2) for randomness in the FSM methodology.
• To assess the reliability of the FSM methodology for identifying fragile sites using chromosomal breakage data.

Main Methods:

• A Monte Carlo simulation procedure was employed.
• 10,000 chromosomal breakage data sets were simulated for varying numbers of breaks (n) and chromosomal band resolutions (k).
• Empirical and nominal alpha levels were compared to assess test reliability.

Main Results:

• Sparse and discrete chromosomal breakage data cause significant discrepancies between empirical and nominal alpha levels for small numbers of breaks (n < 0.5 k).
• With n >= 0.5 k, the Xs2 distribution stabilizes, yielding non-significant differences between empirical and nominal alpha levels.
• These findings are inherent to the data's nature and impact any statistical model for fragile site identification.

Conclusions:

• The FSM methodology requires a minimum of 0.5 mapped chromosomal breaks per individual per band (n >= 0.5 k) for reliable fragile site identification at alpha = 0.05.
• The study provides a critical threshold for data requirements in fragile site analysis.
• Results are applicable to all statistical models analyzing chromosomal breakage data for fragile site identification.