For this study, 150 diarrhea samples were obtained from Maigaiti and Hutubi counties in Xinjiang, China, and categorized them into three samples for viral metagenome sequencing based on the time and place of collection. For clinical validation, 252 clinical samples were collected from different cattle farms in Makit, Hutubi, Urumqi, Changji and Bole Counties, Xinjiang, China. All the samples were collected in the form of anal swabs from calves with clinically suspected diarrhea. Ethical approval for this study was obtained from the Laboratory Animal Welfare and Ethics Committee of Xinjiang Agricultural University (Approval No. 2022038). All samples were immediately homogenized in DEPC-treated, RNase-free phosphate-buffered saline (PBS, pH 7.4) at a 1:5 (w/v) ratio, vortexed at 3,000 rpm for 2 min. and stored at −80°C.

After homogenization in PBS, samples were submitted to Shanghai Tanpu Biotechnology Co., Ltd (Shanghai, China) for viral metagenomic sequencing. Genomic DNA and RNA were extracted using the Viral DNA Kit (Solarbio, China, Cat. No. D3892 and Invitrogen, USA; Cat. No. 15596026) according to the manufacturer’s protocol. Genomic DNA was sheared into 300~500 bp fragments by Covaris M220 ultra-sonicator (peak incident power: 50 W, duty factor: 20%, cycles per burst: 200). RNA was fragmented in Mg²⁺-containing buffer (94°C, 5 min.) and reverse transcribed into cDNA using RT Easy™ II Kit (Foregene, China; Cat. No. RT-01022). DNA and cDNA libraries were prepared with the Illumina TruSeq Nano DNA/RNA Library Prep Kit and quantified via Qubit 4.0. Libraries were sequenced on an Illumina NovaSeq 6000 platform (PE150 mode), which generated approximately 20 Gb raw data per sample. Raw reads were filtered using Fastp v0.20.0 (with the following parameters: remove reads with Q-score < 20, length < 50 bp or adapter contamination), and reads that aligned to the bovine genome (ARS-UCD1.2) were removed using BBmap v38.51. Clean reads were then classified via Kraken2 v2.1.2 with the Standard Database and Bracken v2.6.2 estimated species abundance. Finally, raw viral metagenomic sequencing read archive were submitted to the National Center for Biotechnology Information (NCBI) Sequence Read Archive (SRA) repository under project number PRJNA1244845.

The nucleotide sequences of the RNA-dependent RNA polymerase (RdRp) gene of BNeV (GenBank accession no. MN607031.1), nucleocapsid (N) gene of BCoV(LC494175.1), 5’ untranslated region (5’UTR) of BVDV (AJ133738.1) and viral protein 1 (VP1) gene of BEV (D00214.1) were synthesized by Beijing Tsingke Biotechnology Co. Ltd. (China). Each target sequence was cloned into the pUC57 plasmid vector using EcoR I and Hind III restriction sites ( Supplementary Figure 1S ). The recombinant plasmids were transformed into Escherichia coli DH5α competent cells, purified using the Plasmid Mini Kit (Qiagen, Germany) and quantified via Qubit 4.0 Fluorometer (Thermo Fisher Scientific, USA). Sanger sequencing confirmed 100% identity to the reference sequences. Aliquots were stored at −20°C as quantification standards for downstream assays.

Along with crRNAs targeting protospacer adjacent motif (PAM) sites, polymerase chain reaction (PCR)- and RAA-specific primers were designed based on conserved sequences of the BNeV RdRp, BCoV N, BVDV 5’UTR and BEV VP1 genes ( Table 1 ), and RAA reactions were performed using the RAA Nucleic Acid Amplification Kit (Qitian, China, Cat. No. QT-RAA-100) in a 25.0 μL system containing the following: 12.5 μL reaction buffer, 1.0 μL forward/reverse primers (10.0 μM each), 2.5 μL magnesium acetate (280.0 mM), 1.0 μL DNA template and 7.0 μL nuclease-free ddH₂O. Amplification products were analyzed using 2% agarose gel electrophoresis (100 V, 30 min.) and visualized using a Gel DocTM XR+ System (Bio-Rad, USA). Standard plasmids (1×10⁶ copies/μL) were amplified at 34°C, 37°C, 40°C and 43°C for 30 min. to determine the optimal reaction temperature. Then, under the optimal temperature, reactions were allowed to occur for 10, 15, 20 and 30 min. to identify the minimal required duration. For comparison, conventional RT-PCR assays were established using PrimeSTAR Max DNA Polymerase (Takara, Japan) under the following conditions: 95°C for 5 min.; 35 cycles of 95°C for 30 s, 56–60°C for 30 s, 72°C for 30 s and final extension at 72°C for 5 min.

The CRISPR/Cas12a detection system was established as follows: a 30 μL reaction mixture contained 3 μL 10 × NEB Buffer 2.1 (New England Biolabs, USA; Cat. No. B7202S), 1.5 μL Cas12a nuclease (1 μM; New England Biolabs, USA, Cat. No. M0653T), 1.5 μL crRNA (200 nM; synthesized by Tsingke Biotechnology, China), 1.5 μL fluorescent reporter (FAM-ddT-ssDNA-BHQ1, 500 nM; Tsingke Biotechnology Co., Ltd.), 4 μL RAA amplification product (template) and 18.5 μL nuclease-free ddH₂O. The mixture was incubated at 37°C for 15 min. in a thermal cycler (Bio-Rad, USA) that was protected from light. Fluorescence signals were visualized using a blue LED transilluminator (Azure Biosystems, USA), and they were quantified with ImageJ v1.53 (NIH, USA) ( Figure 1 ). For optimization of Cas12a nuclease concentration, reactions were performed with Cas12a at 100, 200, 400, 600 and 800 nM (crRNA fixed at 200 nM). For optimization of crRNA concentration, crRNA was evaluated at 100–450 nM (in 50 nM increments) with Cas12a fixed at 200 nM. All reactions included negative controls (nuclease-free ddH₂O) and positive controls (1 × 10⁶ copies/μL standard plasmid).

Total RNA was extracted from clinical samples using TRIzol™ Reagent according to the manufacturer’s protocol. RNA concentration and purity were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). RNA (1 μg) was reverse transcribed into cDNA using the PrimeScript™ RT Reagent Kit (Takara, Japan) with random hexamer primers under the following conditions: 37°C for 15 min., 85°C for 5 s. Furthermore, 252 clinical samples collected from Xinjiang, China, were analyzed simultaneously using both the established RAA-CRISPR/Cas12a assay and conventional PCR.

Statistical analyses were performed using SPSS 17.0 software (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation (SD), depending on data normality assessed by the student t-test, and statistical significance was defined as P < 0.05, with asterisks denoting the following: Note: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

The clean reads were assembled de novo using SPAdes (v3.14.1). Contigs ≥1,000 bp were retained for further analysis ( Table 2 ). Assembly statistics revealed 46,176 contigs ≥1,000 bp, including 3,273 (≥5,000 bp), 946 (≥10,000 bp), 151 (≥25,000 bp), and 38 (≥50,000 bp). Viral contigs were annotated using NCBI nt via BLAST+ with an E-value cutoff of ≤1E-5. Across all samples, the following nine viral families were identified: Myoviridae, Reoviridae, Adenoviridae, Coronaviridae, Leviviridae, Podoviridae, Flaviviridae, Caliciviridae and Picornaviridae ( Figures 2A, B ). Furthermore, taxonomic classification resolved the following 11 genera: Rotavirus, Escherichia virus, Enterobacteria phage, Shigella virus, Pestivirus, Nebovirus, Shigella phage, Yersinia virus, Enterovirus, Human mastadenovirus and Betacoronavirus ( Figures 2C, D ).

Using recombinant plasmids of BNeV, BCoV, BVDV and BEV as templates, RAA isothermal amplification generated specific bands of 160 bp (BNeV), 102 bp (BCoV), 197 bp (BVDV) and 192 bp (BEV), which were consistent with expected sizes ( Figures 3A–D ). For temperature optimization, reactions performed at 34–43°C revealed that 37°C yielded the highest amplification efficiency for all four viruses, and their band intensity declined at higher temperatures ( Figures 3E–H ). For time course optimization, detectable amplification occurred at 37°C within 10 min., and fluorescence signals peaked at 20 min. ( Figures 3I–L ). Grayscale analysis showed significant signal enhancement at 20 rather than 15 min. (P < 0.05); however, no such difference was observed between 20 and 30 min. Thus, 20 min. was selected as the optimal reaction time.

To maximize collateral cleavage activity of Cas12a, two candidate crRNAs targeting conserved regions of BNeV and BEV were evaluated. crRNA1 for both viruses exhibited significantly higher fluorescence intensity than crRNA2 (P < 0.01, Student’s t-test; Figures 4A, D ), indicating superior target recognition and trans-cleavage efficiency. Similarly, crRNAs for BCoV and BVDV displayed distinct fluorescence signals compared to negative controls ( Figures 4B, C ), confirming their functionality.

Gradient testing revealed pathogen-specific optimal concentrations as follows: BNeV: 400 nM Cas12a + 250 nM crRNA ( Figure 4E ), BCoV: 200 nM Cas12a + 250 nM crRNA ( Figure 4F ), BVDV: 600 nM Cas12a + 300 nM crRNA ( Figure 4G ) and BEV: 800 nM Cas12a + 200 nM crRNA ( Figure 4H ). These concentrations were chosen for subsequent assays to maximize detection sensitivity.

To evaluate the specificity of the RAA-CRISPR/Cas12a and PCR assays for BNeV, BCoV, BVDV and BEV, each system was evaluated against cDNA/DNA from the target viruses, E. coli, S. enterica and nuclease-free water (negative control). In the RAA-CRISPR/Cas12a system, fluorescence signals were observed exclusively for the target viruses ( Figures 5A–D ). Grayscale analysis (ImageJ) confirmed that fluorescence intensities for BNeV, BCoV, BVDV and BEV were significantly higher than those of non-target pathogens and negative controls (P < 0.001, Student’s t-test; Figures 5A–D ). No cross-reactivity was detected with E. coli, S. enterica or negative controls. In the PCR detection system, agarose gel electrophoresis demonstrated amplification bands exclusively for the target viruses ( Supplementary Figure 2S ), with no detectable products for non-target pathogens or negative controls. These results confirmed that both assays exhibit high specificity for their respective targets, showing no cross-reactivity to phylogenetically unrelated pathogens or bacterial DNA.

The detection limits of the RAA-CRISPR/Cas12a and PCR assays were assessed using serial dilutions that are 10 times of standard plasmids. Fluorescence signals (grayscale analysis, P < 0.01 vs. negative control; Student’s t-test) were detectable at 10 copies/μL in the RAA-CRISPR/Cas12a system for BNeV, BVDV and BEV ( Figures 6A, C, D ). Significant fluorescence (P < 0.001) was observed at a concentration of 1 copy/μL in the RAA-CRISPR/Cas12a system for BCoV ( Figure 6B ). In contrast, the detection limits for conventional PCR were 1 × 10⁵ copies/μL for BNeV and BCoV and 1 × 10³ copies/μL for BVDV and BEV ( Supplementary Figure 3S ). The RAA-CRISPR/Cas12a assay demonstrated sensitivity that was 100–100,000 times greater than conventional PCR, with the most significant improvements observed for BCoV (100,000 times) and BNeV (10,000 times).

To evaluate assay reproducibility, intra- and inter-batch variability were assessed using plasmid standards at concentrations of 5 × 10⁶, 5 × 10⁵ and 5 × 10⁴ copies/μL. Intra-batch reproducibility was assessed through triplicate testing within a single run, while inter-batch reproducibility was evaluated by conducting three independent experiments on different days. For BNeV, the intra-batch CVs ranged from 2.16% to 3.35%, while the inter-batch CVs ranged from 1.41% to 2.06% across different concentrations. BCoV exhibited intra-batch CVs of 0.64–3.48% and inter-batch CVs of 1.41–4.88%. For BVDV, intra-batch and inter-batch coefficients of variation (CVs) were 2.21–2.87% and 2.29–3.40%, respectively. BEV demonstrated intra-batch CVs of 2.47–3.15% and inter-batch CVs of 3.41–3.87% ( Table 3 ). No fluorescence signals were observed in the negative controls. These results confirmed that the RAA-CRISPR/Cas12a system is highly reproducible and stable for all four targets.

The diagnostic performance of RAA-CRISPR/Cas12a was evaluated using 252 clinical anal swab samples and compared to conventional PCR. As shown in Table 4 , the RAA-CRISPR/Cas12a assay demonstrated significantly higher positive detection rates for all four pathogens compared to PCR: BNeV (9.52% vs. 5.95%), BCoV (25.79% vs. 10.32%), BVDV (29.37% vs. 5.95%), and BEV (30.16% vs. 9.92%). Statistical analysis revealed that the RAA-CRISPR/Cas12a system exhibited 1.6 to 4.9 times higher sensitivity than PCR across all targets ( Figures 7A, B , Supplementary Figure 4S ). Notably, all PCR-positive samples (n = 81) were consistently identified by the RAA-CRISPR/Cas12a method, thereby confirming 100% specificity for true positives.