Antibodies currently comprise the predominant treatment modality for a variety of diseases; therefore, optimizing their properties rapidly and efficiently is an indispensable step in antibody-based drug development. Inspired by the great success of artificial intelligence-based algorithms, especially deep learning-based methods in the field of biology, various computational methods have been introduced into antibody optimization to reduce costs and increase the success rate of lead candidate generation and optimization. Herein, we briefly review recent progress in deep learning-based antibody optimization, focusing on the available datasets and algorithm input data types that are crucial for constructing appropriate deep learning models. Furthermore, we discuss the current challenges and potential solutions for the future development of general-purpose deep learning algorithms in antibody optimization.
Deep Learning ; Humans ; Antibodies/chemistry* ; Algorithms ; Artificial Intelligence ; Drug Development

Liver transplantation (LT) has become a standard treatment for end-stage liver diseases, and graft injury is intricately associated with poor prognosis. Granzyme B (GZMB) plays a vital role in natural killer (NK) cell biology, but whether NK-derived GZMB affects graft injury remains elusive. Through the analysis of single-cell RNA-sequencing data obtained from human LT grafts and the isolation of lymphocytes from mouse livers following ischemia-reperfusion injury (IRI), we demonstrated that 2NK cells with high expression of GZMB are enriched in patients and mice. Both systemically and liver-targeted depletion of NK cells led to a notable reduction in GZMB⁺ cell infiltration, subsequently resulting in diminished graft injury. Notably, the reconstitution of Il2rg ^-/- Rag2 ^-/- mice with purified Gzmb-KO NK cells demonstrated superior outcomes compared to those with wild-type NK cells. Crucially, global knockout of GZMB and pharmacological inhibition exhibited remarkable improvements in liver function in both mouse IRI and rat LT models. Moreover, a phosphorylated derivative of FDA-approved vidarabine was identified as an effective inhibitor of mouse GZMB activity by molecular dynamics, which could provide a potential avenue for therapeutic intervention. Therefore, targeting NK cell-derived GZMB during the LT process suggests potential therapeutic strategies to improve post-transplant outcomes.

Folliculogenesis is essential for production of female gametes in vertebrates. However, the molecular mechanisms underlying follicle development, particularly apoptosis regulation in ovary, remain elusive. Here, we generated sox3 knockout zebrafish lines using CRISPR/Cas9. sox3 knockout led to follicle development retardation and a reduced fecundity in females. Comparative analysis of transcriptome between sox3 and wild-type ovaries revealed that Sox3 was involved in pathways of ovarian steroidogenesis and apoptosis. Knockout of sox3 promoted follicle apoptosis and obvious apoptosis signals were detected in somatic cells of stages III and IV follicles of sox3 ovaries. Moreover, Sox3 can bind to and activate the promoter of cyp19a1a. Up-regulation of Cyp19a1a expression promoted 17β-estradiol synthesis, which inhibited apoptosis in follicle development. Thus, Sox3 functions as a regulator of Cyp19a1a expression, via 17β-E2 linking apoptosis suppression, which is implicated in improving female fecundity.

OBJECTIVE To investigate genes associated with aminoglycosides modification enzymes(AMEs) in Pseudomonas aeruginosa(PAE) isolated from ICU patients.METHODS Drug-reisistant genes encoding AMEs such as aac(3)-Ⅰ,aac(3)-Ⅱ,aac(6′)-Ⅰ,aac(6′)-Ⅱ,ant(3″)-Ⅰ and ant(2″)-Ⅰ were detected by polymerase chain reaction(PCR)(amplification) in 21 PAE isolates.RESULTS The positive rates of aac(3)-Ⅱ,aac(6′)-Ⅰ,aac(6′)-Ⅱ,ant(3″)-Ⅰ,ant(2″)-Ⅰ and aac(3)-Ⅰ genes were positive in 19.0%,23.8%,9.5%,4.8%,19.0% and 0% of 21 isolates,respectively. Drug-resistant genes encoding AMEs were detected positively in 42.8% of 21 isolates.(CONCLUSIONS) AMEs genes are present in high percentage of PAE isolated from ICU patients.

OBJECTIVE To investigate genes associated with the drug-resistance of ?-lactam antibiotics in Pseudomonas aeruginosa(PAE) isolated from clinical patients. METHODS ?-Lactamase genes including TEM,SHV,OXA-10,PER,VEB,GES,CARB,IMP,VIM,SPM,GIM,DHA,FOX,MOX and oprD2 were detected by PCR amplification in 33 PAE isolates. RESULTS TEM,SHV,GES,CARB and VIM genes were positive in 100%,6.1%,6.1%,9.1% and 12.1% of 33 isolates,respectively.The deletion of oprD2 gene was found in 22 isolates.Other ?-lactamase genes were absent in all isolates.By PCR amplification,DNA sequencing and BLASTn comparison analysis,the CARB genes of 2 strains were demonstrated to be CARB-3 and the VIM genes of 2 strains were VIM-2. CONCLUSIONS P.aeruginosa carries various beta-lactamase genes in clinical PAE patients,and the deletion ratio of oprD2 gene is high.