Table 1 Association of CoDA features with clinical, molecular, and genetic subgroups are presented

From: Collagen disorder architecture features are associated with clinical, molecular, genetic factors and survival outcomes in colon cancer

Variable type

Variables

CoDA features

p values

Bonferroni-corrected p-values (*p)

CLINICAL

OVERALL STAGE

CF fragmentation

0.001

0.004

CF bundling

0.02

0.10

CF anisotropy

0.004

0.01

CF density

0.02

0.12

CF rigidity

0.01

0.04

T STAGE

CF fragmentation

0.01

0.07

CF bundling

0.01

0.05

CF anisotropy

0.007

0.03

CF density

0.005

0.02

CF rigidity

0.009

0.04

N STAGE

CF fragmentation

0.008

0.04

CF bundling

0.001

0.007

CF anisotropy

0.001

0.003

CF density

0.006

0.03

CF rigidity

0.007

0.03

M STAGE

CF fragmentation

0.006

0.03

CF bundling

0.004

0.02

CF anisotropy

0.01

0.05

CF density

0.02

0.13

CF rigidity

0.01

0.09

MOLECULAR

CMS1

CF fragmentation

0.001

0.006

CF bundling

0.01

0.05

CF anisotropy

0.001

0.007

CF density

0.01

0.09

CF rigidity

0.02

0.13

CMS2

CF fragmentation

0.02

0.13

CF bundling

0.007

0.03

CF anisotropy

0.01

0.05

CF density

0.01

0.07

CF rigidity

0.02

0.11

CMS3

CF fragmentation

0.02

0.10

CF bundling

0.006

0.02

CF anisotropy

0.02

0.12

CF density

0.009

0.04

CF rigidity

0.009

0.04

CMS4

CF fragmentation

0.01

0.05

CF bundling

0.02

0.12

CF anisotropy

0.009

0.04

CF density

0.01

0.05

CF rigidity

0.02

0.12

GENETIC

KRAS

CF fragmentation

0.006

0.02

CF bundling

0.001

0.007

CF anisotropy

0.01

0.07

CF density

0.002

0.01

CF rigidity

0.003

0.01

BRAF

CF fragmentation

0.03

0.15

CF bundling

0.005

0.02

CF anisotropy

0.001

0.04

CF density

0.007

0.03

CF rigidity

0.00

0.002

NRAS

CF fragmentation

1.00

1.00

CF bundling

0.01

0.07

CF anisotropy

1.00

1.00

CF density

0.03

0.15

CF rigidity

0.008

0.04

  1. The Mann–Whitney U test was used to assess differences in CoDA features: CF fragmentation, CF bundling, CF anisotropy, CF density, and CF rigidity, across various clinical stages, CMS subtypes, and mutation-defined groups. Bonferroni correction was applied to adjust for multiple comparisons, with a significance threshold of 0.05 for both raw and corrected p-values (denoted as p and *p, respectively). Statistically significant associations for Bonferroni corrected p values are shown in bold. Significant associations were observed across several variables, particularly for N stage, M stage, and CMS3, where all five CoDA features remained significant after correction (*p < 0.05). Similarly, CoDA features were strongly associated with KRAS mutation status, while NRAS mutation showed minimal association except for CF rigidity. These findings highlight that distinct collagen architectural patterns, as captured by CoDA features, are significantly associated with key clinical stages, molecular subtypes, and genetic mutations in CC, reflecting their potential to capture tumor heterogeneity across multiple biological axes.