Update publication. To construct the spectral library, 100 μg peptides from each sample were mixed and separated into 10 fractions by a RIGOL L‐3000 HPLC (Puyuanjingdian science and technology Ltd., Beijing, China). In detail, the mixture peptides were injected into a Gemini-NX C18 110A column (Size 250×4.6 mm, Particle size 5μm, Phenomenex, Germany) at 1mL/minute using mobile phase A (2% ACN, pH=10) and B (98% ACN, pH=10) with the following gradient: 0–25 minutes, 5%–30% B. The separated fractions were loaded into an analytical column (15 cm, ID 150 μm, Particle size 1.9 μm) using an EASY-nano-LC system at a flow rate of 600nL/min. The separation gradient was as follows: The gradient was set as follows: 10–14% solvent B for 12 min, 14–26% solvent B for 45 min, 26–42% solvent B for 10 min, 42–95% solvent B for 1 min, 95% solvent B for 7 min. MS1 scans was acquired from 350 to 1400 m/z with the resolution of 60,000. The top twenty precursor ions were selected for MS2 by HCD fragmentation at normalized collision energy (NCE) of 28 with a resolution of 15,000. The automatic gain control (AGC) was set to 3e6 for full MS1 and 5e4 for MS2, with maximum ion injection times of 80 and 45 ms, respectively. The DIA was performed the same LC system condition mass spectrometer as above.data-independent acquisition (DIA) database construction analysis. The full scan was set at a resolution of 60,000 at a range of 350 to 1400 m/z; DIA scans parameter was set with resolution 30,000; NCE: 28%; AGC target: 3e6 and maximal injection time: auto. Forty-five variable DIA windows were set for DIA acquisition. Spectronaut pulsar X 12.0 (Biognosys, Schlieren, Switzerland) was used for identification and quantification. For identification, the DDA raw files were searched against the Aeromonas hydrophila fasta database (4,121 entries, downloaded on July 15, 2019 from uniprot) to generate spectral library using BGS factory setting. Peptides FDR\ PSMs FDR\ Proteins FDR were all set as 1%, Best 3-6 fragments per peptides were chosen for spectral library. The iRT Calibration R square was set as 0.8. For quantification, DIA data were input into the software. The iRT regression type was set as Local (Non- Linear) Regression. Every peptide contains at least 3 fragment-ions. All results were filtered by a Q value cutoff of 0.01 (corresponds to a FDR of 1%). P-value estimator was performed by Kernel Density Estimator.