Introduction

YTH N⁶-methyladenosine RNA binding protein 1 (YTHDF1) is a N⁶-methyladenosine (m⁶A) reader that is overexpressed in colorectal cancer (CRC), and it induces colorectal tumorigenesis and metastatic progression. Nevertheless, no pharmacological approach is available for targeting YTHDF1.

Aims & Methods

To this end, we performed Epi-Drug CRISPR/Cas9 library screening for druggable targets that impair the oncogenic function of YTHDF1 in CRC.
Epi-Drug CRISPR/Cas9 library screen was used to identify genes selectively depleted in CRC cells overexpressing YTHDF1. Clinical significance of gene candidates was evaluated in independent CRC patient cohorts. Interplay between gene candidates and YTHDF1 using integrative m⁶A sequencing, RNA sequencing, Ribosome-sequencing, ribosome profiling analyses, and liquid chromatography-mass spectrometry. Targeting YTHDF1 druggable targets were assessed by pharmacological inhibitors and vesicle-like nanoparticles (VNPs)-encapsulated siRNA.

Results

Using Epi-Drug CRISPR/Cas9 library screen, we identified RUVBL1 and RUVBL2 as top essential druggable target genes in CRC cells overexpressing YTHDF1. RUVBL1/2 are overexpressed in primary CRCs from two independent cohorts (Beijing and TCGA) and they predict poor survival in patients with CRC. Functionally, silencing RUVBL1 or RUVBL2 suppressed tumorigenesis of YTHDF1-overexpressing cell lines, primary CRC organoids and CRC xenografts in nude mice. Mechanistically, YTHDF1 directly binds to m⁶A-modified RUVBL1/2, as determined by integrated m6A sequencing, RNA sequencing, ribosome sequencing, leading to increased RUVBL1/2protein translation. Interestingly, co-immunoprecipitation and mass spectrometry revealed that RUVBL1/2 proteins directly interact with YTHDF1 and its associated translation initiation factors. Moreover, RUVBL1/2 functions as the essential components of YTHDF1 translational complex, and their depletion abrogated YTHDF1-induced protein translation and bio-synthesis. As a consequence, RUVBL1/2 knockout largely abolished YTHDF1-induced oncogenic signaling, especially the activation of MAPK and PI3K signaling pathways. To verify RUVBL1/2 as potential therapeutic targets in YTHDF1-expressing CRC, we next employed pharmacological RUVBL1/2 inhibitors and nanoparticle-encapsulated siRUVBL1/2 to target RUVBL1/2 in CRC models. In support of our notion, targeting RUVBL1/2 with pharmacological inhibitors or in vivo siRNA delivery suppressed the growth of YTDHF1-overexpressing CRC cell in vitro and regressed tumor xenografts in vivo.

Conclusion

Our study identified RUVBL1/2 as druggable target genes of YTHDF1. Targeting RUVBL1/2 is a promising strategy for treatment of CRC with high YTHDF1 expression.

Introduction

The integration of artificial intelligence (AI) into nursing practice offers a significant opportunity to enhance patient care and support clinical decision-making. This study examines the deployment of a customized generative language model (GPT) as a clinical assistant in outpatient digestive care. The tool aims to streamline access to evidence-based information and promote structured patient follow-up. While its implementation holds considerable potential for improving efficiency and fostering clinical learning, it also presents challenges related to validation, user adoption, and building trust among nursing professionals.

Results & Findings

A digital tool named Digestive Nursing Follow-Up was developed and piloted in the Gastroenterology Departments of two university hospitals in Spain during Q1 2025. The tool was integrated into four clinical domains: inflammatory bowel disease, gastroesophageal reflux disease, irritable bowel syndrome, and advanced endoscopy. The AI assistant was designed to: Collect and organize structured clinical data; Provide evidence-based clinical suggestions aligned with institutional protocols; Facilitate learning and care continuity. Prior to implementation, participating nurses underwent a one-hour training session. The tool was then incorporated into routine consultations, and nurses were surveyed pre- and post-intervention using a 5-point Likert Scale. Open-ended feedback was also collected.

Summary & Discussion

80% of nurses rated the tool as easy to use following training, and 75% reported high satisfaction, citing improved time management and a reduction in repetitive tasks. Positive perceptions centered on enhanced consultation efficiency and clinical safety. However, concerns were expressed regarding the balance between technology and patient-centered care, as well as the need for ongoing training and more personalized outputs.

Conclusion

Among the strengths identified were improved time management, support for clinical decision-making, and enhancement of standardized nursing practice. The main weaknesses were resistance to change and the initial learning curve, which required additional support. Opportunities include the potential of AI to strengthen professional training, enhance professional autonomy, and contribute to the standardization of clinical procedures. The tool reinforces clinical safety by being based on established protocols, and its use can accelerate their implementation and promote continuous learning, helping to consolidate a more efficient care model. The observed technological dependence was more associated with increased consultation efficiency, highlighting the importance of progressive and balanced integration with clinical practice.

References

1. Exploring the deep learning of artificial intelligence in nursing: a concept analysis with Walker and Avant’s approach. BMC Nurs. 2024;23:170. doi:10.1186/s12912-024-02170-x.
2. Examining the Role of AI in Changing the Role of Nurses in Patient Care: Systematic Review. JMIR Nurs. 2025;1:e63335. doi:10.2196/63335.
3. Artificial intelligence in healthcare: a scoping review of perceived threats to patient rights and safety. Arch Public Health. 2024;82:1414. doi:10.1186/s13690-024-01414-1.
4. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Med Educ. 2023;23:4698. doi:10.1186/s12909-023-04698-z.
5. Utilization of Generative Artificial Intelligence in Nursing Education: A Topic Modeling Analysis. Educ Sci. 2023;14:1234. doi:10.3390/educsci14111234.
6. Human-centered design and evaluation of AI-empowered clinical decision support systems: a systematic review. Front Comput Sci. 2023;5:1187299. doi:10.3389/fcomp.2023.1187299.
7. Generative AI in healthcare: an implementation science informed translational path on application, integration and governance. Implement Sci. 2024;13:357. doi:10.1186/s13012-024-01357-9.
8. Telehealth and Remote Patient Monitoring Innovations in Nursing Practice: State of the Science. OJIN: The Online Journal of Issues in Nursing. 2022;27(2):1-15.
9. Artificial Intelligence and Digital Innovations in Nursing Care. Nurs Rep. 2023;13(1):45-60.

Introduction

Computer-aided diagnosis (CAD) systems are widely implemented in the medical field for the analysis of medical images. PolyDeep is a CAD system that has demonstrated its efficacy in lesion detection.

Aims & Methods

The aim of this study is to determine whether PolyDeep increases the adenoma detection rate (ADR) in the context of expert endoscopists participating in a regional, population-based colorectal cancer screening program in Spain. PolyDeep Advance 3 (NCT05513261) is a multicentric, randomized controlled clinical trial. For the primary endpoint, we aimed to assess whether the ADR is superior in PolyDeep-assisted colonoscopy compared to conventional colonoscopy. As secondary endpoints, we aimed to determine whether PolyDeep-assisted colonoscopy improves the polyp detection rate (PDR), serrated lesion detection rate (SLDR), advanced serrated lesion detection rate (ASLDR), and advanced adenoma detection rate (AADR). We included participants who underwent colonoscopy for screening (i.e. those with a positive faecal immunochemical occult blood test – FIT+) or for surveillance after resection of advanced adenomas. We compared both groups using the chi-square and t-Student test.

Results

Between November 2023 and May 2024, we recruited 857 patients, of whom 30 were excluded. The baseline characteristics of the 827 patients included in the analysis were similar (male: 59.3%, screening colonoscopy: 66.7%, mean age: 61.7 ± 6.5 years, withdrawal time: 10:49 ± 04:32 minutes, colorectal cancer incidence: 2.8%). However, there were significant differences between both groups in the Boston Bowel Preparation Scale (conventional colonoscopy: 7.7 ± 1.3 vs. PolyDeep-assisted colonoscopy: 7.4 ± 1.3, p = 0.01). Regarding ADR, we did not find statistically significant differences between conventional and PolyDeep-assisted colonoscopy (57.7% vs. 60.8%, p = 0.4). Similarly, for the secondary endpoints, there were no significant differences in PDR (68.7% vs. 69.6%, p = 0.8), SLDR (32.0% vs. 29.4%, p = 0.5), ASLDR (5.3% vs. 5.1%, p = 1.0), or AADR (24.5% vs. 23.6%, p = 0.8). In Table 1, we described lesion detection rates between both groups by indication (i.e., FIT+ or surveillance).

Conclusion

In the context of expert endoscopists PolyDeep did not increase the ADR.