Structure-Based Optimization of Pyrazinamide-Containing Macrocyclic Derivatives as Fms-like Tyrosine Kinase 3 (FLT3) Inhibitors to Overcome Clinical Mutations.

Secondary mutations in Fms-like tyrosine kinase 3-tyrosine kinase domain (FLT3-TKD) (e.g., D835Y and F691L) have become a major on-target resistance mechanism of FLT3 inhibitors, which present a significant clinical challenge. To date, no effective drugs have been approved to simultaneously overcome clinical resistance caused by these two mutants. Thus, a series of pyrazinamide macrocyclic compounds were first designed and evaluated to overcome the secondary mutations of FLT3. The representative 8v exhibited potent inhibitory activities against FLT3D835Y and FLT3D835Y/F691L with IC50 values of 1.5 and 9.7 nM, respectively. 8v also strongly suppressed the proliferation against Ba/F3 cells transfected with FLT3-ITD, FLT3-ITD-D835Y, FLT3-ITD-F691L, FLT3-ITD-D835Y-F691L, and MV4-11 acute myeloid leukemia (AML) cell lines with IC50 values of 12.2, 10.5, 24.6, 16.9, and 6.8 nM, respectively. Furthermore, 8v demonstrated ideal anticancer efficacy in a Ba/F3-FLT3-ITD-D835Y xenograft model. The results suggested that 8v can serve as a promising macrocycle-based FLT3 inhibitor for the treatment of AML.

Relationship between the cGAS-STING and NF-κB pathways-role in neurotoxicity.

Neurotoxicity can cause a range of symptoms and disorders in humans, including neurodegenerative diseases, neurodevelopmental disorders, nerve conduction abnormalities, neuroinflammation, autoimmune disorders, and cognitive deficits. The cyclic guanosine-adenosine synthase (cGAS)-stimulator of interferon genes (STING) pathway and NF-κB pathway are two important signaling pathways involved in the innate immune response. The cGAS-STING pathway is activated by the recognition of intracellular DNA, which triggers the production of type I interferons and pro-inflammatory cytokines, such as tumor necrosis factor, IL-1ß, and IL-6. These cytokines play a role in oxidative stress and mitochondrial dysfunction in neurons. The NF-κB pathway is activated by various stimuli, such as bacterial lipopolysaccharide, viral particle components, and neurotoxins. NF-κB activation may lead to the production of pro-inflammatory cytokines, which promote neuroinflammation and cause neuronal damage. A potential interaction exists between the cGAS-STING and NF-κB pathways, and NF-κB activation blocks STING degradation by inhibiting microtubule-mediated STING transport. This review examines the progress of research on the roles of these pathways in neurotoxicity and their interrelationships. Understanding the mechanisms of these pathways will provide valuable therapeutic insights for preventing and controlling neurotoxicity.

Kinase Inhibition via Small Molecule-Induced Intramolecular Protein Cross-Linking.

Remarkable progress has been made in the development of cysteine-targeted covalent inhibitors. In kinase drug discovery, covalent inhibitors capable of targeting other nucleophilic residues (i.e. lysine, or K) has emerged in recent years. Besides a highly conserved catalytic lysine, almost all human protein kinases possess an equally conserved glutamate/aspartate (e.g. E/D) that forms a K-E/D salt bridge within the enzyme active-site. Electrophilic ynamides were previously used as effective peptide coupling reagents and to develop E/D-targeting covalent protein inhibitors/probes. In the present study, we report the first ynamide-based small-molecule inhibitors capable of inducing intramolecular cross-linking of various protein kinases, leading to subsequent irreversible inhibition of kinase activity. Our strategy took advantage of the close distance between the highly conserved catalytic K and E/D residues in a targeted kinase, thus providing a conceptually general approach to achieve irreversible kinase inhibition with high specificity and desirable cellular potency. Finally, this ynamide-facilitated, ligand-induced mechanism leading to intramolecular kinase cross-linking and inhibition was unequivocally established by using recombinant ABL kinase as a representative.

Deep learning-based automated high-accuracy location and identification of fresh vertebral compression fractures from spinal radiographs: a multicenter cohort study.

Background: Digital radiography (DR) is a common and widely available examination. However, spinal DR cannot detect bone marrow edema, therefore, determining vertebral compression fractures (VCFs), especially fresh VCFs, remains challenging for clinicians. Methods: We trained, validated, and externally tested the deep residual network (DRN) model that automated the detection and identification of fresh VCFs from spinal DR images. A total of 1,747 participants from five institutions were enrolled in this study and divided into the training cohort, validation cohort and external test cohorts (YHDH and BMUH cohorts). We evaluated the performance of DRN model based on the area under the receiver operating characteristic curve (AUC), feature attention maps, sensitivity, specificity, and accuracy. We compared it with five other deep learning models and validated and tested the model internally and externally and explored whether it remains highly accurate for an external test cohort. In addition, the influence of old VCFs on the performance of the DRN model was assessed. Results: The AUC was 0.99, 0.89, and 0.88 in the validation, YHDH, and BMUH cohorts, respectively, for the DRN model for detecting and discriminating fresh VCFs. The accuracies were 81.45% and 72.90%, sensitivities were 84.75% and 91.43%, and specificities were 80.25% and 63.89% in the YHDH and BMUH cohorts, respectively. The DRN model generated correct activation on the fresh VCFs and accurate peak responses on the area of the target vertebral body parts and demonstrated better feature representation learning and classification performance. The AUC was 0.90 (95% confidence interval [CI] 0.84-0.95) and 0.84 (95% CI 0.72-0.93) in the non-old VCFs and old VCFs groups, respectively, in the YHDH cohort (p = 0.067). The AUC was 0.89 (95% CI 0.84-0.94) and 0.85 (95% CI 0.72-0.95) in the non-old VCFs and old VCFs groups, respectively, in the BMUH cohort (p = 0.051). Conclusion: In present study, we developed the DRN model for automated diagnosis and identification of fresh VCFs from spinal DR images. The DRN model can provide interpretable attention maps to support the excellent prediction results, which is the key that most clinicians care about when using the model to assist decision-making.

Continuous blood exchange in rats as a novel approach for experimental investigation.

Blood exchange therapy, specifically Whole blood exchange (WBE), is increasingly being utilized in clinical settings to effectively treat a range of diseases. Consequently, there is an urgent requirement to establish convenient and clinically applicable animal models that can facilitate the exploration of blood exchange therapy mechanisms. Our study conducted continuous WBE in rats through femoral and tail vein catheterization using dual-directional syringe pumps. To demonstrate the applicability of continuous WBE, drug-induced hemolytic anemia (DIHA) was induced through phenylhydrazine hydrochloride (PHZ) injection. Notability, the rats of DIHA + WBE group all survived and recovered within the subsequent period. After the implementation of continuous WBE therapy day (Day 1), the DIHA + WBE group exhibited a statistically significant increase in red blood cells (RBC) (P = 0.0343) and hemoglobin (HGB) levels (P = 0.0090) compared to DIHA group. The rats in the DIHA + WBE group exhibited a faster recovery rate compared to the DIHA group, indicating the successful establishment of a continuous blood exchange protocol. This experimental approach demonstrates not just promising efficacy in the treatment of DIHA and offers a valuable tool for investigating the underlying mechanisms of blood exchange. Furthermore, it has a great potential to the advancement of biomedical research such as drug delivery exploration.

Nitrogen-Rich Triazine-Based Covalent Organic Frameworks as Efficient Visible Light Photocatalysts for Hydrogen Peroxide Production.

Covalent organic frameworks (COFs) have been widely used in photocatalytic hydrogen peroxide (H2O2) production due to their favorable band structure and excellent light absorption. Due to the rapid recombination rate of charge carriers, however, their applications are mainly restricted. This study presents the design and development of two highly conjugated triazine-based COFs (TBP-COF and TTP-COF) and evaluates their photocatalytic H2O2 production performance. The nitrogen-rich structures and high degrees of conjugation of TBP-COF and TTP-COF facilitate improved light absorption, promote O2 adsorption, enhance their redox power, and enable the efficient separation and transfer of photogenerated charge carriers. There is thus an increase in the photocatalytic activity for the production of H2O2. When exposed to 10 W LED visible light irradiation at a wavelength of 420 nm, the pyridine-based TTP-COF produced 4244 µmol h-1 g-1 of H2O2 from pure water in the absence of a sacrificial agent. Compared to TBP-COF (1882 µmol h-1 g-1), which has a similar structure but lacks pyridine sites, TTP-COF demonstrated nearly 2.5 times greater efficiency. Furthermore, it exhibited superior performance compared to most previously published nonmetal COF-based photocatalysts.

GCMSFormer: A Fully Automatic Method for the Resolution of Overlapping Peaks in Gas Chromatography-Mass Spectrometry.

Gas chromatography-mass spectrometry (GC-MS) is one of the most important instruments for analyzing volatile organic compounds. However, the complexity of real samples and the limitations of chromatographic separation capabilities lead to coeluting compounds without ideal separation. In this study, a Transformer-based automatic resolution method (GCMSFormer) is proposed to resolve mass spectra from GC-MS peaks in an end-to-end manner, predicting the mass spectra of components directly from the raw overlapping peaks data. Furthermore, orthogonal projection resolution (OPR) was integrated into GCMSFormer to resolve minor components. The GCMSFormer model was trained, validated, and tested using 100,000 augmented data. It achieves 99.88% of the bilingual evaluation understudy (BLEU) value on the test set, significantly higher than the 97.68% BLEU value of the baseline sequence-to-sequence model long short-term memory (LSTM). GCMSFormer was also compared with two nondeep learning resolution tools (MZmine and AMDIS) and two deep learning resolution tools (PARAFAC2 with DL and MSHub/GNPS) on a real plant essential oil GC-MS data set. Their resolution results were compared on evaluation metrics, including the number of compounds resolved, mass spectral match score, correlation coefficient, explained variance, and resolution speed. The results demonstrate that GCMSFormer has better resolution performance, higher automation, and faster resolution speed. In summary, GCMSFormer is an end-to-end, fast, fully automatic, and accurate method for analyzing GC-MS data of complex samples.

Insights into the key odorants in fresh and dried Amomum tsaoko using the sensomics approach.

To identify the key odorants in Amomum tsaoko (AT), volatiles in fresh AT (FAT) and dried AT (DAT) were investigated using molecular sensory science. In addition to this, the sensomics approach was used to confirm the presence of the compound in FAT that contributed the most to its aroma profile. A total of 49 odor-active compounds (43 in FAT and 42 in DAT) with flavor dilution (FD) factors ranging from 1 to 6561 were identified, with eucalyptol exhibiting the highest FD factor of 6561. Odorants with FD factors ≥ 27 were quantitated, and 23 and 20 compounds in FAT and DAT, respectively, with odor activity value ≥ 1 were determined as key odorants. Recombination and omission experiment further indicated that (E)-2-dodecenal, geranial, octanal, (E)-2-octenal, (E)-2-decenal, and eucalyptol contributed significantly to the overall aroma profile of FAT. After drying of FAT, the concentrations of aldehydes decreased significantly, whereas those of terpene hydrocarbons increased. Multivariate statistical analysis revealed that 26 FAT and 23 DAT odorants were biomarker compounds.

Quanzhen Yiqi decoction attenuates inflammation in mice with smoking-induced COPD by activating the Nrf2/HO-1 pathway and inhibiting the NLRP3 inflammasome.

OBJECTIVE: Quanzhen Yiqi decoction (QZYQ) is a traditional Chinese medicine for treating chronic obstructive pulmonary disease. METHODS: Mice were exposed to cigarette smoke (CS) 6 days/week (40 cigarettes/day) for 24 weeks and then intragastrically administered QZYQ (4.72, 9.45, or 18.89 g/kg) or dexamethasone (DEX, 0.6 mg/kg) for 6 weeks. We examined the lung function and collected bronchoalveolar lavage fluid for inflammatory cell and cytokine quantification. The pathological lung changes, ROS and oxidative biomarkers were measured. We used immunohistochemistry and western blotting to evaluate the levels of Nrf2/HO-1, NLRP3/ASC/Caspase1/IL-1ß/IL-18. RESULTS: The CS group showed significant increases in the forced vital capacity, lung resistance, and chord compliance and a lower FEV50/FVC compared with the control, and QZYQ improved these changes. In addition, QZYQ effectively reduced emphysema, immune cell infiltration, and airway remodeling. QZYQ stimulated HO-1 expression and reduced oxidative stress through the Nrf2 pathway. QZYQ inhibited the production of NLRP3/ASC/Caspase-1 to inhibit IL-1ß and IL-18. CONCLUSION: Our study suggested that QZYQ can improve the function and histology of the lungs and reduce inflammatory cell recruitment. QZYQ inhibits ROS production and NLRP3 inflammasome activation by upregulating Nrf2 to reduce lung injury. The anti-inflammatory effects of QZYQ are similar to those of DEX.

Effects of Dietary Inosine 5'-Monophosphate Supplementation on the Growth Performance and Salinity and Oxidative Stress Resistance of Gibel Carp (Carassius auratus gibelio).

An 88-day feeding trial was conducted to evaluate the effects of dietary inosine 5'-monophosphate (5'-IMP) on the growth performance and salinity and oxidative stress resistance in the juvenile gibel carp CAS III (Carassius auratus gibelio; initial body weight: 7.48 g). Four isonitrogenous and isoenergetic diets containing exogenous 5'-IMP were formulated. P1, P2, P3 and P4 were diets containing 5'-IMP at four concentrations (0, 1, 2 and 4 g kg-1). The four diets were randomly allotted to triplicate tanks in a recirculating system. After the feeding trial, six fish per tank were netted randomly and placed into 12‱ saline water to test their response to salinity stress. The results indicated that the feed conversion rate was enhanced by dietary supplementation with 5'-IMP. The appetite, plasma neuropeptide Y level and feeding rate of the P3 group were lower than those in the control treatment group. Dietary supplementation with 5'-IMP improved the osmoregulatory adaptation of gibel carp under acute salinity stress. Six hours after the salinity stress treatment, in the dietary 5'-IMP treatment group, the plasma cortisol and K+ concentrations were lower and the Na+/K+-ATPase activity was greater than that in the control group. Dietary supplementation with 5'-IMP promoted the expression of the glucocorticoid receptors NKA-α1b and NKCC and retarded the expression of Hsp70 in P4-treated gill filaments and kidneys. Dietary supplementation with 5'-IMP resulted in a stable oxidative-stress-resistant phenotype characterized by increased levels of cellular antioxidants, including SOD, catalase, glutathione peroxidase, glutathione reductase and MPO. The above results of the current study demonstrate that supplementation of 5'-IMP can promote feed utilization and have positive influences on the salinity and oxidative stress resistance of gibel carp.

Plant-Derived Caffeic Acid and Its Derivatives: An Overview of Their NMR Data and Biosynthetic Pathways.

In recent years, caffeic acid and its derivatives have received increasing attention due to their obvious physiological activities and wide distribution in nature. In this paper, to clarify the status of research on plant-derived caffeic acid and its derivatives, nuclear magnetic resonance spectroscopy data and possible biosynthetic pathways of these compounds were collected from scientific databases (SciFinder, PubMed and China Knowledge). According to different types of substituents, 17 caffeic acid and its derivatives can be divided into the following classes: caffeoyl ester derivatives, caffeyltartaric acid, caffeic acid amide derivatives, caffeoyl shikimic acid, caffeoyl quinic acid, caffeoyl danshens and caffeoyl glycoside. Generalization of their 13C-NMR and 1H-NMR data revealed that acylation with caffeic acid to form esters involves acylation shifts, which increase the chemical shift values of the corresponding carbons and decrease the chemical shift values of the corresponding carbons of caffeoyl. Once the hydroxyl group is ester, the hydrogen signal connected to the same carbon shifts to the low field (1.1~1.6). The biosynthetic pathways were summarized, and it was found that caffeic acid and its derivatives are first synthesized in plants through the shikimic acid pathway, in which phenylalanine is deaminated to cinnamic acid and then transformed into caffeic acid and its derivatives. The purpose of this review is to provide a reference for further research on the rapid structural identification and biofabrication of caffeic acid and its derivatives.

Mechanically stimulated osteocytes maintain tumor dormancy in bone metastasis of non-small cell lung cancer by releasing small extracellular vesicles.

Although preclinical and clinical studies have shown that exercise can inhibit bone metastasis progression, the mechanism remains poorly understood. Here, we found that non-small cell lung cancer (NSCLC) cells adjacent to bone tissue had a much lower proliferative capacity than the surrounding tumor cells in patients and mice. Subsequently, it was demonstrated that osteocytes, sensing mechanical stimulation generated by exercise, inhibit NSCLC cell proliferation and sustain the dormancy thereof by releasing small extracellular vesicles with tumor suppressor micro-RNAs, such as miR-99b-3p. Furthermore, we evaluated the effects of mechanical loading and treadmill exercise on the bone metastasis progression of NSCLC in mice. As expected, mechanical loading of the tibia inhibited the bone metastasis progression of NSCLC. Notably, bone metastasis progression of NSCLC was inhibited by moderate exercise, and combinations with zoledronic acid had additive effects. Moreover, exercise preconditioning effectively suppressed bone metastasis progression. This study significantly advances the understanding of the mechanism underlying exercise-afforded protection against bone metastasis progression.

Characterization of a novel anti-PVRIG antibody with Fc-competent function that exerts strong antitumor effects via NK activation in preclinical models.

Poliovirus receptor-related immunoglobulin domain-containing protein, or PVRIG, is a newly discovered immune checkpoint that has emerged as a promising target for cancer immunotherapy. It is primarily expressed on activated T and natural killer (NK) cells, and once engaged with its ligand, PVRL2, it induces inhibitory signaling in T cells, thereby promoting the functional exhaustion of tumor-infiltrating lymphocytes (TILs). Here, we characterized IBI352g4a, a novel humanized anti-PVRIG antibody with Fc-competent function, explored the mechanism of its antitumor activity in preclinical models, and systemically evaluated the contribution of FcrR engagement to PVRIG blockade-induced antitumor activity. IBI352g4a binds to the extracellular domain of human PVRIG with high affinity (Kd = 0.53 nM) and specificity, and fully blocks the interaction between PVRIG and its ligand PVRL2. Unlike other immune checkpoints, IBI352g4a significantly induced NK cell activation and degranulation, but had a minimal effect on T-cell activation in in vitro functional assays. IBI352g4a induced strong antitumor effect in several preclinic models, through in vivo mechanism analysis we found that both NK and T cells contribute to the antitumor effect, but NK cells play predominant roles. Specifically, a single dose of IBI352g4a induced significant NK cell activation in TILs, but T-cell activation was observed only after the second dose. Moreover, the Fc effector function is critical for both NK cell activation and treatment efficacy in vitro and in vivo. Our study, for the first time, demonstrates that both NK activation and FcrR engagement are required for antitumor efficacy induced by PVRIG blockade.

Effect and mechanism of Prunella vulgaris L. extract on alleviating lipopolysaccharide-induced acute mastitis in protecting the blood-milk barrier and reducing inflammation.

ETHNOPHARMACOLOGICAL RELEVANCE: According to ancient literature, Prunella vulgaris L. (P vulgaris) alleviates mastitis and has been used in China for many years; however, there are no relevant reports that confirm this or the mechanism of its efficacy. AIM OF THE STUDY: To explore the anti-acute mastitis effect and potential mechanism of P vulgaris extract. MATERIALS AND METHODS: First, the active ingredients and targets of P vulgaris against mastitis were predicted using network pharmacology. Next, the relevant active ingredients were enriched using macroporous resins and verified using UV and UPLC-Q-TOF-MS/MS. Lastly, a mouse model of acute mastitis was established by injecting lipopolysaccharides into the mammary gland and administering P vulgaris extract by oral gavage. The pathological changes in mammary tissue were observed by HE staining. Serum and tissue inflammatory factors were measured by ELISA method. MPO activity in mammary tissue was measured using colorimetry and MPO expression was detected by immunohistochemistry. The expression of tight junction proteins (ZO-1, claudin-3, and occludin) in mammary tissue was detected by immunofluorescence and Western blot. iNOS and COX-2 in mammary tissue were detected by Western blot. MAPK pathway and NF-κB pathway related proteins were also detected by Western blot. RESULTS: Network pharmacology predicted that phenolic acids and flavonoids in P vulgaris had anti-mastitis effects. The contents of total flavonoids and total phenolic acids in P vulgaris extract were 64.5% and 29.4%, respectively. UPLC-Q-TOF-MS/MS confirmed that P vulgaris extract contained phenolic acids and flavonoids. The results of animal experiments showed that P vulgaris extract reduced lipopolysaccharide-induced inflammatory edema, inflammatory cell infiltration, and interstitial congestion of mammary tissue. It also reduced the levels of serum and tissue inflammatory factors TNF-α, IL-6, and IL-1ß, and inhibited the activation of MPO. Furthermore, it downregulated the expression of MAPK and NF-κB pathway-related proteins. The expressions of ZO-1, occludin, and claudin-3 in mammary gland tissues were upregulated. CONCLUSIONS: P vulgaris extract can maintain the integrity of mammary connective tissue and reduce its inflammatory response to prevent acute mastitis. Its mechanism probably involves regulating NF-κB and MAPK pathways.

Mechanism of Action of "Astragali Radix-Cassia Twig-Poria" in Chronic Heart Failure: A Network Pharmacology Approach.

This study utilizes network pharmacology analysis to investigate the components, targets, and pathways involved in the treatment of chronic heart failure (CHF) with the combination of "Astragali Radix-Cassia Twig-Poria." The TCMSP, GeneCards, OMIM, PharmGkb, TTD, and DrugBank databases were utilized to identify the active ingredients and targets of this combination for CHF. Protein interactions were derived from the STRING database, and Cytoscape was used to construct the "drug-component-target-disease" network and protein interactions network. The GO function and KEGG signaling pathway were enriched, and molecular docking was performed to verify the stability of the core components and their targets. The study identified 41 active ingredients, 101 targets (including 94 related to CHF), 9 core targets, and 26 core ingredients of "Astragali Radix-Cassia Twig-Poria." Additionally, 1444 GO entries and 140 KEGG pathways (including 36 related to CHF) were found. Molecular docking results confirmed the binding ability of the combination to core targets. Overall, this study provides valuable insights into the key components, targets, and pathways involved in the treatment of CHF with "Astragali Radix-Cassia Twig-Poria," contributing to further research on its pharmacological effects.

Bacterial host adaptation through sequence and structural variations of a single type III effector gene.

Molecular mechanisms underlying quantitative variations of pathogenicity remain elusive. Here, we identified the Xanthomonas campestris XopJ6 effector that triggers disease resistance in cauliflower and Arabidopsis thaliana. XopJ6 is a close homolog of the Ralstoniapseudosolanacearum PopP2 YopJ family acetyltransferase. XopJ6 is recognized by the RRS1-R/RPS4 NLR pair that integrates a WRKY decoy domain mimicking effector targets. We identified a XopJ6 natural variant carrying a single residue substitution in XopJ6 WRKY-binding site that disrupts interaction with WRKY proteins. This mutation allows XopJ6 to evade immune perception while retaining some XopJ6 virulence functions. Interestingly, xopJ6 resides in a Tn3-family transposon likely contributing to xopJ6 copy number variation (CNV). Using synthetic biology, we demonstrate that xopJ6 CNV tunes pathogen virulence on Arabidopsis through gene dosage-mediated modulation of xopJ6 expression. Together, our findings highlight how sequence and structural genetic variations restricted at a particular effector gene contribute to bacterial host adaptation.

Case study on the influence of serving temperature on the aroma release and perception of Huangjiu, a fermented alcoholic beverage.

Serving temperature plays a crucial role in influencing the sensory experience of consumers. In this context, this study investigated the influence of serving temperature on the aroma release and perception of a typical fermented alcoholic beverage named Huangjiu. A quantitative sensory description analysis was conducted, determining serving temperature significantly influenced the 17 sensory attributes in both semi-dry and semi-sweet Huangjiu. The variation in the contents of 41 volatiles in the Huangjiu with temperature was investigated using gas chromatography-ion mobility spectrometry, resulting in volatile content significantly increasing above 30 ℃. The partial least squares discriminant analysis was conducted to predict the variable importance for the projection (VIP) of volatiles, and 22 volatiles (VIP > 1) were screened. These 22 volatiles were confirmed as key odorants influenced by serving temperature though aroma addition experiments. The findings would provide a reference for the effects of serving temperature on the flavor perception of fermented alcoholic beverages.

Large-Scale Glass-Transition Temperature Prediction with an Equivariant Neural Network for Screening Polymers.

The practically infinite chemical and morphological space of polymers makes them pervasive with applications in materials science but challenges the rational discovery of new materials with favorable properties. Polymer informatics aims to accelerate materials design through property prediction and large-scale virtual screening. In this study, a new method (Lieconv-Tg) has been developed to predict glass-transition temperature (Tg) values from repeating units of polymers based on Lieconv, which is equivariant with transformations from any specified Lie group. The introduction of equivariance allows the prediction of molecular properties from their 3D structures, independent of orientation and position. A total of 27,659 homopolymers with Tg values were collected from PolyInfo, and a standard data set containing 7166 polymers (named data set_Tg) was created for training a robust Lieconv-Tg model. Using the 3D coordinates as input, Lieconv-Tg performs better than Edge-Conditioned Convolution (ECC), and the mean absolute error (MAE) is significantly reduced by ∼6 from ∼30 to ∼24 on both the validation set and the test set, and the R2 value for both the validation set and the test set can reach 0.90. Lieconv-Tg is thus used to screen promising candidates from a benchmark database named PI1M with 995,800 generated polymers. However, there are some implausible repeating units in PI1M. To get more reasonable candidates from PI1M, a new filtering method has been accomplished by utilizing Morgan fingerprints at the polymerization points (MF@PP) of repeating units in data set_Tg. The combination of a standard data set, Lieconv-Tg, and a more reasonable screening strategy provides new directions in materials design for polymers.