Published September 15, 2026 | Version v1
Event Open

Quantitative Mapping of Data-Derived Biochemical State Transitions under Chemical Perturbation in Strawberry and Blackberry

Description

Abstract
Chemical perturbations can alter enzymatic function not only in magnitude but also in the organization of concentration-dependent response trajectories. Here, we developed Biochemical State-Transition Analysis (BSTA), a data-driven framework for quantifying biochemical response architecture from concentration-dependent ascorbate peroxidase (APX) activity in strawberry and blackberry. The concentration-dependent dataset comprised 504 experimentally measured APX observations across 16 fruit × treatment profiles, supplemented by 48 baseline measurements obtained from the same profiles. Profile-specific responses were transformed into a multidimensional functional-state representation and analyzed using Gaussian mixture modeling to identify data-derived response states. Concentration-dependent trajectories were subsequently characterized using complementary descriptors of transition position, transition sharpness, state persistence (Pstate), Response-change entropy (Hresponse), and trajectory similarity. Replicate-aware bootstrap resampling was used to quantify uncertainty in concentration-dependent transition descriptors.
The analysis identified two data-derived response states, and all 100 computational bootstrap refits of the selected GMM configuration converged successfully, supporting optimization stability rather than biological state reproducibility. Across the 16 fruit × treatment profiles, each reconstructed trajectory exhibited a single concentration-ordered transition from the higher-Rz state to the lower-Rz state, whereas transition position, sharpness, persistence, and functional retention varied among treatments and fruit matrices. A thresholded response-profile similarity network contained 59 edges among 120 possible profile pairs, indicating that 59 of 120 profile pairs met the prespecified threshold for normalized trajectory-shape similarity. Fruit-specific differences were strongly dependent on the chemical perturbation, with a pronounced fruit × treatment interaction in integrated concentration-dependent response magnitude. The results demonstrate that biochemical response architecture cannot be adequately represented by a single endpoint or midpoint parameter. BSTA provides a complementary phenotype-level framework for quantifying how biochemical responses are organized across controlled chemical perturbation gradients without assigning mechanistic meaning to statistically derived states or network relationships.

Files

BSTA_FINAL_FIGURES_ALL FFFinal.zip

Files (19.8 MB)

Name Size
md5:29f79eedb302960107f5e070c8fb8c40
13.7 MB Preview Download
md5:df6018a3ec13897b093d7aedbc9b10b8
31.3 kB Download
md5:db613d2cee5c364f69f925c43927a7d5
355.1 kB Preview Download
md5:df2edf3a4d325323d08c1a407a281be2
5.4 MB Preview Download
md5:ff977e0f603e6db1730b2ddc16bfcca7
6.5 kB Download
md5:1905ac3469a1ae993c1bcb54afbf53b1
5.7 kB Download
md5:4e89f5dbfbaeb479da31de362c06e398
11.7 kB Download
md5:857ba9fc2601529132b2db12bcb72d92
12.6 kB Download
md5:e6c3999155119d805bcec442b7951d35
20.2 kB Download
md5:f8831645cccd30582d66ab04f98dcdbd
15.2 kB Download
md5:87c01238c0f6c4a6676c45ec75b476bd
9.5 kB Download
md5:d7ff15b53ebd4e6f75fed3453e5a191a
21.4 kB Download
md5:d7ff15b53ebd4e6f75fed3453e5a191a
21.4 kB Download
md5:d7ff15b53ebd4e6f75fed3453e5a191a
21.4 kB Download
md5:7749fc0baf4925a6f9d962ed0a24a8df
6.6 kB Download
md5:2fcff7cb62fdf6fd6bf767dda31c7a04
6.9 kB Preview Download
md5:0e5384f342a6e19c8b25ed76cf1b1060
8.8 kB Download
md5:db60033b4a493abc30807ecf5688848c
9.0 kB Download
md5:34f72d012b89f053d92e8723a3310531
34.7 kB Download
md5:83b218998b06d10641e16c2dfb6605bc
15.6 kB Download
md5:a7543ee897dd5c3c44f32355f63b408c
31.9 kB Download