Explore chapters and articles related to this topic
Statistics for Genomics
Published in Altuna Akalin, Computational Genomics with R, 2020
If you have observed the table output of the summary() function, you must have noticed there are some other outputs, such as “Residual standard error”, “Multiple R-squared” and “F-statistic”. These are metrics that are useful for assessing the accuracy of the model. We will explain them one by one.
Toxicology Studies
Published in Nusrat Rabbee, Biomarker Analysis in Clinical Trials with R, 2020
We know that the omnibus F statistics from classic ANOVA does not distinguish between the means of treatment groups. So, contrast coding is used to compare means of treatment groups to one or more control groups.
F
Published in Filomena Pereira-Maxwell, Medical Statistics, 2018
Also known as variance-ratio test. A significance test that is carried out to compare the variances of two independent groups with respect to a Normally distributed variable, i.e. to test the assumption of homoscedasticity. The test statistic, with degrees of freedom equal to each respective sample size, is calculated as the ratio of the larger to the smaller variance (see below). Another common application of the F-test is in comparing between-groups to within-groups variability when an analysis of variance is carried out to compare the means of several groups (Box O.3, p. 246). Under the null hypothesis of no difference between the groups, these two components of variability are the same and their ratio (F-statistic) is equal to one. The F distribution (which is followed by the F-statistic when the null hypothesis is true) has two different degrees of freedom: the number of groups minus 1 (between-groups), and the total number of observations minus the number of groups (within-groups). When comparing two independent groups, the F-test yields the same P-value as the unpaired t-test. An overall F-test is used in least squares regression to test the joint significance of all variables in a model. Nested models, where one model is an extension of the other, may be compared using a partial F-test. See also Bartlett’s test, likelihood ratio test, stepwise regression, residual F-test.
The Effects of Fatty Acids on Primary Liver Cancer: A Two-Sample Mendelian Randomization Study
Published in Nutrition and Cancer, 2023
Yuan Liu, Jian He, Longjiao Cai, Aimin Leng
SNPs selected as IVs for FAs represent the global human genetic variation (30). First, SNPs associated with six FAs with a genome-wide significance threshold of p < 5 × 10−8 were identified as prospective IVs to identify relevant SNPs that meet MR specifications. Additionally, we clumped and discarded SNPs at linkage disequilibrium (LD) threshold r2 < 0.001 and a window size of 1000 kb. The F statistic and R2 statistic for SNPs were determined to set the statistical strength of IVs. The formula for calculating the F statistic is as follows: N denotes the number of samples and K is the number of IVs (31). Here, R2 can show the degree to which genetic IVs explain the exposure factors (32). Weak IVs were identified in SNPs with F < 10 and were excluded (31).
4-Formylphenyl boronic acid grafted amino MCM-41 for efficient adsorption of Cu(II) ions in aqueous medium: isotherm, kinetic and optimization studies
Published in Toxin Reviews, 2022
Inderpreet Kaur, Navjot Kaur, Bhupinder Pal Singh, Rajeev Kumar, Jyoti Chawla
The graphical illustration of the response surface of the adsorbed Cu(II) ions relative to three variables pH, Concentration, and temperature was used to understand the interactions between variables and validate their optimal level to a highest removal percentage. The optimal values were obtained from solution of regression equation and by analysis of response surface plots. It is important to highlight that response surface studies depict the optimum conditions for maximum removal of Cu(II) ions from aqueous solution by using b-MCM within specific range of pH 4–5.38, concentration 11.87–81.70 mg/L and temperature 20.79–38.81 °C for 89.01–91.72 removal percentage corresponding to adsorbent dose 1 g/L. The adequacy of the models was vindicated by the analysis of variance (ANOVA). The ANOVA for percentage removal of Cu(II) is given in Table 2. The model F-values (ratio of mean square for individual term to the mean square for residual) implies that the model is significant. The Probability of f-statistics value (Prob > F value) is used to test the null hypothesis. Experimental and model predicted values of the response variable (Percentage removal) were compared after optimization of variables. The plot between experimental (actual) and predicted values for percentage removal of Cu(II) is also shown in Figure 4. It can be seen from Figure 4 that both the values were in logical agreement with each other. It indicates the good correlation between input and response variables.
For whom the circadian clock ticks? Investigation of PERIOD and CLOCK gene variants in bipolar disorder
Published in Chronobiology International, 2021
Zeynep Yegin, Gokhan Sarisoy, Ayse Erguner Aral, Haydar Koc
SPSS 22 statistics software was used for all statistical analyses performed in this study. Descriptive statistics are given as n (%) and mean ± standard deviation for categorical and numerical variables, respectively, in both control and patient groups. The association between the genotypes of all genes that were used in our study and BD risk were determined by odds ratios (ORs) and 95% confidence intervals (CIs) using logistic regression model. Furthermore, we used χ2 or Fisher’s exact test to evaluate the relationship between genotypes of circadian genes and the clinical characteristics. In addition, the F statistics was used to compare the average scores of behavioral variables between genotypes. Tukey multiple comparison test was used to find means that are significantly different from each other. p < .05 value was used to determine the statistical significance for all tests conducted. Statistical power analysis was applied to decide how many observations to work with for this study. Power values for different sample numbers were calculated with G-Power software version 3.1.9.4 for power analysis, and the power of the statistical test for a total of 242 observations, 121 for each of the control and patient groups, was calculated as 99.9% at the 0.05 significance level.