Table 1 Top single-omic and multi-omic analytes for predicting disease survival in PDAC in the MT-Pilot cohort

From: The Molecular Twin artificial-intelligence platform integrates multi-omic data to predict outcomes for pancreatic adenocarcinoma patients

Analytes

No. of samples

No. of features

TP

FP

TN

FN

ACC (95% CI)

PPV (95% CI)

Sensitivity

Specificity

Multi-omic

39

6,363

26

4

7

2

0.85 (0.73–0.96)

0.87 (0.75–0.99)

0.93

0.64

Plasma proteins

51

257

32

8

6

5

0.75 (0.63–0.86)

0.80 (0.68–0.92)

0.86

0.43

RNA fusions

57

29

35

12

8

2

0.75 (0.64–0.87)

0.74 (0.62–0.87)

0.95

0.40

Tissue proteins

49

1,130

32

10

4

3

0.73 (0.61–0.86)

0.76 (0.63–0.89)

0.91

0.29

Plasma lipids

51

406

34

12

2

3

0.71 (0.58–0.83)

0.74 (0.61–0.87)

0.92

0.14

Clinical and surgical pathology

74

331

47

19

5

3

0.70 (0.60–0.81)

0.71 (0.60–0.82)

0.94

0.21

RNA gene expressions

57

2,000

33

14

6

4

0.68 (0.56–0.80)

0.70 (0.57–0.83)

0.89

0.30

Computational pathology

71

819

34

11

13

13

0.66 (0.55–0.77)

0.76 (0.63–0.88)

0.72

0.54

DNA CNVs

72

648

43

20

4

5

0.65 (0.54–0.76)

0.68 (0.57–0.80)

0.90

0.17

DNA INDELs

72

126

39

17

7

9

0.64 (0.53–0.75)

0.70 (0.58–0.82)

0.81

0.29

DNA SNVs

72

611

45

23

1

3

0.64 (0.53–0.75)

0.66 (0.55–0.77)

0.94

0.04

CA 19-9 presurgery

63

1

17

15

20

11

0.59 (0.47–0.71)

0.53 (0.40–0.65)

0.61

0.57

  1. Single and multi-omic analytes predicting DS are listed in decreasing order of predictive performance for DS, arranged by accuracy and PPV. For each analyte, the of number of samples available and features extracted for that respective analyte are shown. The predictive performance for each analyte is based on the best-performing model. TP, true positive; FP, false positive; TN, true negative; FN, false negative; ACC, accuracy. The bold ACC and PPV values indicates the best-performing analyte.