For TCs which are judged to be landing on Hainan Island, PRESS and its progressive optimal algorithm and MLRM
can be used to forecast TCs' characteristic factors (including latitude, longitude, the lowest center pressure, and wind speed).
Ten SNPs with the lowest unadjusted p-values for each of the three NKAs were tested, along with corresponding significant covariates, to estimate their effect and relative contributions to the measured level of each NKA using MLRM. In the final models, six, five, and five SNPs were retained as significant predictors for 1NKA, 2NKA, and TNKA, respectively (Table 1).
Of the 16 SNPs highly associated with NKA levels identified by GWA and MLRM analyses, 4 were associated with the same sequence, leaving 12 SNPs that were unique genic or intergenic sequences (Table 2).
By ranking the 10 SNPs with the lowest p-values for each MCA biomarker in PLINK, we were able to construct final exposure models and investigate the contributions of multiple SNPs using MLRM. Furthermore, analysis of predicted networks may minimize false-positive findings without missing important pathways due to a mandated stringent threshold.
The MLRM results showed that ethnicity was significant only for 2NKA and TNKA levels but not for the 1NKA level.
In the models presented here, the number of parameters is likely to be large relative to the population sample size because we used the 10 SNPs with the lowest p-value for each intermediate phenotype (NKA level) to establish the final MLRM. Therefore, the [R.sup.2] of the model may be increased.
To accomplish this, we developed multiple linear regression models (MLRMs) using SAS, which included dermal naphthalene level and significant personal and workplace factors ([alpha] = 0.1).