Table 1 Resource requirements to generate LD matrices for marginal analysis

From: Computationally efficient meta-analysis of gene-based tests using summary statistics in large-scale genetic studies

Sample size

CPU time (min)

Memory (Gb)

Storage (Mb)

REMETA

MetaSTAAR

REMETA

MetaSTAAR

REMETA

MetaSTAAR

25,000

0.95

23.55

1.34

2.77

1.16

2.67

50,000

2.93

29.81

2.26

5.41

1.41

3.10

100,000

9.15

41.15

3.95

12.33

1.73

3.86

200,000

28.06

75.16

7.17

30.63

2.16

4.93

  1. Per gene LD matrices were computed on chr. 20 using exome variants in subsets of UKB. Note that MetaSTAAR requires an LD matrix per trait in each study while REMETA requires only a single LD matrix per study. Experiments were run using a single core.
Back to article page