Hematologic malignancies remain among the most difficult cancers to treat, challenged by profound heterogeneity, treatment-induced immune dysfunction, and the frequent emergence of drug resistance. Beyond tumor-intrinsic mechanisms, dysbiosis of the gut microbiome is increasingly recognized as a critical determinant of therapeutic outcomes, shaping hematopoiesis, immune responses, and drug metabolism. Bacteriophage (phage) therapy has re-emerged as a precision tool capable of selectively eradicating pathogenic taxa while preserving commensal short-chain fatty acid-producing communities. Preclinical and early human studies demonstrate that phages can recalibrate microbial ecosystems, disrupt antibiotic-tolerant biofilms, and enrich metabolites such as butyrate that support mucosal integrity and immune balance. Mechanistically, phage DNA enriched with CpG motifs engages Toll-like receptor 9, activating dendritic cells and enhancing cytotoxic T lymphocyte responses, suggesting dual benefits in infection control and anti-tumor immunity. Emerging applications extend further, with engineered phages serving as vectors for CRISPR-Cas gene editing, targeted cytokine delivery, and nanocarrier platforms for leukemia therapy. Despite translational promise, major hurdles persist, including immunogenicity, horizontal gene transfer, resistance evolution, and regulatory uncertainty. Addressing these challenges through GMP-compliant manufacturing, metagenomics-guided personalization, and AI-optimized cocktail design could establish phage therapy as a microbiome-informed adjunct to overcome drug resistance in blood cancers. However, direct clinical evidence of phage therapy efficacy in hematologic malignancies remains limited, and current data are largely derived from preclinical and compassionate-use contexts.

On event-based social platforms such as Meetup and Douban, online groups serve as more than virtual communities for users to share experiences, they also provide an essential pathway for users to discover and participate in offline events. As the number of groups grows, it imposes the need of the study of online group recommendation. Despite there being many existing approaches to solve this problem, they all ignore the phenomenon that the groups that users participate in often contain a number of similar users. This phenomenon implies that similar users play a crucial role in identifying the groups that users are likely to join. In order to exploit similar users to improve the recommendation performance, we propose an effective multi-scale contrastive learning system for online Group Recommendation services, which is with a two-Tower model in event-based social networks (Tower4GR). Specifically, we first adopt the two-tower model to capture the interactive signals within the sequences and groups. We then incorporate the features of similar users into the sequence encoder, and aggregate the relevant users' features into the group encoder, through which the preferred groups of similar users are more likely to be discovered by the target user. Finally, we propose an effective multi-scale contrastive learning framework for the two-tower architecture. It derives self-supervision signals from both same-scale data and cross-scale data, thereby extracting more meaningful data patterns. Moreover, the framework strengthens the cooperative associations between two towers. Extensive experiments on three real-world datasets from Meetup demonstrate the superiority of our proposed model over existing state-of-the-art models.