### Table 2: True complexes associated to MaWish re ned alignments.

2007

"... In PAGE 11: ... and 3. in both Table2 and 3 have equal hypergeometric score, showing that the coverage, that is, number of proteins of an alignment contained in its best true MIPS module, does not change. Alignment 2.... In PAGE 11: ... Alignment 2. in Table2 covers 50% of the true complex, while its re nement in Table 3 covers the entire true complex (Casein kinase II, consisting of 4 proteins).... ..."

Cited by 1

### Table 7.18. True class values rates delivered by the association rules model

2007

### Table 1). Note that the x-axis aligns with east axis, the y-axis aligns with north, and z points upward; a yaw angle of zero degrees is associated with true north.

"... In PAGE 3: ... Table1 . Sensors used to obtain the robot states.... ..."

Cited by 1

### Table 2. The power comparisons of the three tests: F-HPM and QPDT (after Bonferroni correction) for the associations and QST for linkage between the seven candidate genes and the five quantitative traits at statistical significance level 5%. There are three true gene-trait associations and the power for these three pairs is denoted in bold face font.

2002

"... In PAGE 8: ... For each gene and each replication, we simulate 200 samples to evaluate the statistical significance of the observed test statistic. The power comparisons between the F-HPM method and two other methods, the QPDT [2] and QST [12] (a score test of linkage for quantitative traits using haplotypes in extended pedigrees and using hierarchical clustering method to group the haplotypes into two groups), based on these 50 replications are summarized in Table2 . Because Q3 and Q4 have no associations with any of the genes, they are not shown in Table 2.... In PAGE 8: ... According to Table 1, G1, G3, G4, G5, and G7 have no association with all the quantitative traits, G2 is associated with Q5 only and G6 is associated with Q1 and Q2. It can be seen from Table2 that the false positive rate of our method is within the 95% confidence interval of the nominal level, i.... ..."

Cited by 7

### Table 2. The power comparisons of the three tests: F-HPM and QPDT (after Bonferroni correction) for the associations and QST for linkage between the seven candidate genes and the five quantitative traits at statistical significance level 5%. There are three true gene-trait associations and the power for these three pairs is denoted in bold face font.

"... In PAGE 8: ... For each gene and each replication, we simulate 200 samples to evaluate the statistical significance of the observed test statistic. The power comparisons between the F-HPM method and two other methods, the QPDT [2] and QST [12] (a score test of linkage for quantitative traits using haplotypes in extended pedigrees and using hierarchical clustering method to group the haplotypes into two groups), based on these 50 replications are summarized in Table2 . Because Q3 and Q4 have no associations with any of the genes, they are not shown in Table 2.... In PAGE 8: ... According to Table 1, G1, G3, G4, G5, and G7 have no association with all the quantitative traits, G2 is associated with Q5 only and G6 is associated with Q1 and Q2. It can be seen from Table2 that the false positive rate of our method is within the 95% confidence interval of the nominal level, i.... ..."

### Table 3: Con dence value associated with the true class for every image in the database using edge direction coherence vector-based 5-NN classi er; 89% of the 2; 716 images were correctly classi ed with a con dence of 75% and above.

1998

Cited by 29

### Table 1 gives the parameters of the posterior pdf. In this table, k is the hypothesis number, k is the hypoth- esis, p( kjZ1,Z0) is the posterior probability of hypothesis k, ^ x11 ( k) and P1( k) are the posterior mean and covariance of p(x11jZ1,Z0, k). Note that since the covariance matrix is diagonal, only the variances in the x and y directions have been given. The components have been sorted in descending order of posterior probability, so 1 has the largest posterior probability. The ten components with largest posterior probabil- ity, along with the true association, are shown in table 1. Although the true hypothesis is at position 18, the most probable component is in error only for target 6 (which is incorrectly hypothesised to be undetected). The required mixture pdf is given by

1999

"... In PAGE 12: ...005 5.185 Overall : x11( ) = (2:403; 6:279)T P1( ) = 14:860 0:391 0:391 8:138 True values : x11 = (5:250; 7:012)T; B = (5:000; 5:000)T Table1 : Top 10 components of the posterior pdf of target 1 and true association. Table 1 gives the parameters of the posterior pdf.... ..."

Cited by 1

### Table 2. Equation for predicting ratio of observed to true association parameter in logistic regression (b*/b) in a semi-ecological study when between-worker (within-group) variance is directly proportionalto group mean (s2 B = cmg), but within-workervariance (sW) is homoscedastic; 207 simulations each repeated 1000 times (n = 207 000); five group means {1.1, 2.1, 3.1, 4.1, 5.1}, two repeats per subject; residual between-simulation variance: 0.0007; residual within-simulation variance: 0.0266 Predictor Range in simulation Effect estimate Standard error P (effect = 0)

2005

"... In PAGE 6: ... The metamodel for bias in logistic regres- sion seems to account for the observed patterns in simulation. The final metamodel parameters for logistic regres- sion are shown in Table2 . They include the same interaction terms as the metamodel for bias in Cox proportional-hazards regression and similar effects of sample size and within-worker variability.... In PAGE 6: ... The main striking difference is that bias appears to be independ- ent of the association parameter except through its interaction with the heteroscedasticity parameter. (NB: The estimate of the effect of the association parameter per se is not shown in Table2 , because it was not significant in the final metamodel for logistic regression, bb = 0.007, standard error (bb = 0.... In PAGE 6: ...02, P = 0.7, while other estimates were the same as shown in Table2 .) In the metamodel of bias in logistic regression, the heteroscedasticity parameter itself has the opposite effect than in Cox proportional- hazards models: to create positive bias.... In PAGE 6: ...egression, when (for example) b=0.4, c = 0.1, k = 50 and sW = 0.5, we simply plug in values into the regression equation described in Table2 (model input in bold font): b*/b=0.... ..."

### Table 1. Equation for predicting ratio of observed to true association parameter in Cox proportional-hazards regression (a*/a) in a semi-ecological study when between-worker (within-group) variance is directly proportional to group mean (s2 B = cmg), but within-worker variance (sW) is homoscedastic; 207 simulations each repeated 1000 times (n = 207 000); five group means {1.1, 2.1, 3.1, 4.1, 5.1}, two repeats per subject; residual between-simulation variance: 0.0005; residual within-simulation variance: 0.0074

2005

"... In PAGE 5: ... The metamodel for bias in Cox proportional- hazards regression seems to account for the observed patterns in simulation. The final metamodel parameters for Cox proportional-hazards regression are presented in Table1 . The cubic polynomials of the heteroscedasti- city parameter were retained in the final model because they produced an improvement in model fit compared with having a quadratic term of the heteroscedasticity parameter.... In PAGE 5: ...hen (for example) a=0.4, c = 0.1, k = 50 and sW = 0.5, we simply plug in values into the regression equation described in Table1 (model input in bold font): a*/a=1.... ..."

### Table 3: Precision and recall in percent for a manually annotated set of 1 500 sentences. Filters are applied successively as described in the algorithm and im- plementation section. A true positive is defined as the correct identification of an association between an enzyme class and a disease-related concept, indepen- dently of the number of correctly identified instances of that association.

2005