Skip to main content

Table 2 Differentially abundant features identified by different methods based on the lung cancer exosomal lipids data

From: SDA: a semi-parametric differential abundance analysis method for metabolomics and proteomics data

Feature ID

\(\hat {\gamma }_{g}\)

\(\hat {\beta }_{g}\)

q SDA

q 2 T

q 2 W

q ELRT

C47H86O6

0.56

-1.17

0.02

0.01

0.25

—

C53H94O6

1.97

-0.7

0.02

0.02

0.08

—

C57H108O6*

1.13

-0.89

0.02

0.18

0.3

0.33

C59H104O6

2.54

-0.23

0.02

0.03

0.08

—

C54H100O6

1.3

-0.57

0.04

0.07

0.14

—

C49H92O6*

1.3

-0.66

0.05

0.26

0.32

0.33

C39H79N2O6P1*

—

0.38

0.07

0.7

0.74

0.73

C40H80N1O8P1*

—

0.31

0.07

0.38

0.32

0.33

C51H94O6*

1.87

-0.48

0.07

0.26

0.32

0.33

C52H98O6*

0.59

-0.8

0.07

0.18

0.32

—

C56H104O6*

0.99

-0.57

0.07

0.13

0.25

—

C56H106O6

-0.3

-0.94

0.07

0.04

0.3

—

C59H106O6*

1.03

-0.7

0.07

0.17

0.25

—

C59H112O6

-0.49

-0.91

0.07

0.01

0.13

—

C56H102O6*

1.13

-0.54

0.08

0.18

0.3

—

  1. FDR threshold was 0.1. Estimations of γ and β as well as q-values from different methods are presented. Lipid assignments of those features are provided in Table S1 in Additional file 2. * indicates features only identified by SDA. — indicates results not available. For C39H79N2O6P1 and C40H80N1O8P1, the calculation of \(\hat {\gamma }_{g}\) is not available because there is no zero value in the cancer samples. For the ELRT method, q-values for many features were not available