“borderline significant”

Of the previously identified risk loci, we replicated 19 at the genome-wide significance level, and seven were borderline significant ( P < 10 −6 , Table S2 ).

Borderline significance (5 × 10 −8 < p < 1 × 10 −6 ) for association with systolic BP was achieved for four index SNPs at four separate loci; there were four SNPs showing borderline significant associations with diastolic BP at four separate loci.

Given that the statistical power in the top1 datasets was relatively low due to the small sample size (4017 individuals), we investigated the p-values of the two eGFRcrecys-specific signals, and found that one (rs9532933 on chromosome 13) was borderline significant for the other two traits ( P = 1.1 × 10 −6 for creatinine and 3.0 × 10 −7 for cystatin C, Table S7 ), while the other (rs116542696 on chromosome 20) was borderline significant only for cystatin C ( P = 0.20 for creatinine and 1.3 × 10 −7 for cystatin C, Fig. 5d–f , Table S7 ), suggesting that these signals are likely not eGFRcrecys-specific, but shared with at least one for the two biomarkers.

However, this SNP was only borderline significant in the Norwegian sample (p-value = 1.8 × 10 − 6 ).

Validation of genetic variants Validation of all nine potentially deleterious genetic variants detected in whole genome sequencing data in 102 SPAID-affected and 62 SPAID-unaffected Shar-Pei using Kompetitive Allele Specific PCR (KASP) and gel electrophoresis revealed a borderline significant P-value for TGFBR3 :g.57204844A > G located in TGFBR3 (transforming growth factor beta receptor 3; P = 0.050), but a highly significant P-value for MTBP :g.19383758G > A (P = 2.664E-06) located in MTBP (Mdm2, transformed 3 T3 cell double minute 2, p53 binding protein ).

Finally, rs1451385 is a borderline significant eQTL (p = 6.7 × 10 −6 ) in subcutaneous adipose tissue in GTEx ( www.gtexportal.org/home ) for AC003090.1 (ENSG00000223561.2), a long intergenic noncoding RNA (lincRNA) with unknown function, located 100 kb from rs1451385.

The effect of IL-5 changed from borderline significant to significantly negative (p < 0.00001) and the heterogeneity decreased (I 2 = 49%) ( Figure 9 ).

In addition, twelve of the significant SNPs in the NHGRI Family Cohort were borderline significant in the meta-analysis ( P < 1 × 10 − 5 ) (Table 3 , Figure S2), of which eight had homogeneous effects in both cohorts ( P het > 0.1).

Across 16 novel and replicable loci, we found that multi-ancestry SLE-only GWAS already yielded genome-wide significant P values in 3 loci and borderline significant P values for the remaining 13 loci (highest P value = 8.19 × 10 −5 ).

The following variables were significant or borderline significant (BS): maternal age ( P =BS), parity ( P <.0001), height ( P <.01), smoking (<0.05), and induction of labor ( P <.0001).

Two CpG-sites were borderline significant in HD – cg18222192 (MIR708)( p < 10E-05, p FDR = 5.81E-02) and cg01299774 (MIR4456)(p < 10E-06, p FDR = 5.81E-02).

Differences were considered as ‘not significant’ in cases with p-values > 0.1, as ‘borderline significant’ ( # ) with p-values between 0.05 and 0.1, as ‘significant’ (*) with p values between 0.01 and 0.05, as ‘highly significant’ (**) with p values between 0.001 and 0.01, as ‘very highly significant’ (***) with p values between 0.0001 and 0.001 and as ‘extremely significant’ (****) with p-values < 0.0001.

Of the four SNPs with significant or borderline significant IL-2 associations in the whole blood assay, only rs10507173 was associated with CD4+ IL-2 production ( p <0.0001, general linearized model), though the association was driven by a single minor allele homozygote outlier (2H) and was not significant under a dominant model ( p = 0.95).

Relative to central hospitals, patients at hospitals in the most peripheral quartile had significantly higher odds of death, but patients at hospitals in quartile 2 were no different from the reference, and those in quartile 3 were borderline significant for reduced mortality (OR (95% CI): Q1 2.10 (1.76–2.51), p<0.0001; Q2 1.00 (0.87–1.15), p = 0.997; Q3 0.89 (0.79–1.00), p = 0.050; Q4 (reference)).

Among them, brain natriuretic peptide (BNP) was the only biomarker with a borderline significant association with IMI (HR per doubling of BNP level (95% confidence interval (95% CI)) = 1.33 (1.15, 1.55), P = 1.63 × 10 −4 , FDR = 0.11 in the discovery sample and 1.40 (1.00, 1.94), P = 0.049 in the validation sample; Extended Data Fig. 2 ).

The rs748096191 genotype also showed significant association with LDL, and non-HDL cholesterol levels and borderline significant association with total cholesterol level after Bonferroni correction ( p = 2.00 × 10 −4 , p = 2.68 × 10 −4 , and p = 8.20 × 10 −4 , respectively).

Three additional associations were borderline significant after multiple testing: a PRS for high blood pressure was associated with renal cell carcinoma (OR = 1.20, p = 2.10 × 10 –4 ), a PRS for deep vein thrombosis had the most significant association with pancreatic cancer (OR = 1.13, p = 1.10 × 10 –4 ), and a PRS for peripheral artery disease was associated with lower risk for glioma (OR = 0.90, p = 2.10 × 10 –5 ; note because other cancers were used as a control this could alternatively be interpreted as high risk for non-glioma cancers.

This SNP was statistically significantly associated with risk of CRC in our unconditional analysis ( P = 1.2×10 −4 ) but was borderline significant with respect to our pre-specified threshold in an analysis conditional on the region’s lead index-correlated variant, rs12474044 ( P = 2.6×10 −4 ).

Hazard ratios of revision surgery were higher with lower final score or less improvement in Mayo hip score at 2-years and borderline significant/non-significant at 5-years, respectively: (1) score ≤55 with hazard ratios of 2.24 (95 % CI, 1.45, 3.46; p = 0.0003) and 1.70 (95 % CI, 1.00, 2.92; p = 0.05) of implant revision subsequently, compared to 72-80 points; (2) no improvement or worsening score with hazard ratios 3.94 (95 % CI, 1.50, 10.30; p = 0.005) and 2.72 (95 % CI, 0.85,8.70; p = 0.09), compared to improvement >50-points.

The predicted change of binding affinity of ESRRA was borderline significant ( P = 0.0004), however since ESRRA is expressed in ovarian tissue and has been shown to bind the SF1 TFBS ( Bonnelye et al. , 1997 ), we decided to include this TF in the subsequent analyses.

Significant heterogeneity was found in the pooled analysis of need for ETI and borderline significant heterogeneity was found for mortality (Cochran's Q chi-square test, p = 0.0004; I 2 = 70.1% for intubation and Cochran's Q chi-square test, p = 0.060; I 2 = 44.1% for mortality).

However, a borderline significant result was found for EHMT2 (p=0.0005 for progression-free survival after BCG induction and maintenance).

Nevertheless, the recording-first group showed quantitatively greater improvement while improvement in the quote-first group was only borderline significant (recording-first improvement = 1.29, t 51 = 3.67, p < 0.0006, d = 0.49; quote-first improvement = 0.59, t 43 = 1.96, p < 0.057, d = 0.29; Figure 4C ).

This difference was significant for A-07 and U-25 tumors ( P = .035 and P = .026, respectively; Figure 3 C ), borderline significant for D-12 and R-18 tumors ( P = .13 and P = .14, respectively; Figure 3 C ), and highly significant when A-07, U-25, D-12, and R-18 tumors were pooled together ( P < .001; Figure 3 D ).

A borderline significant inverse association was observed while comparing parity with nulliparous, with summarized RR = 0.79 (95% CI: 0.60–1.06; I 2 = 90.9%, P < 0.001).