Biology subjects

Lin, C.

Publications and source records attributed to Lin, C..

At least 55 records · Page 3Linked to original sources

bioRxiv2025

Leveraging cell-type specificity and similarity improves single-cell eQTL fine-mapping

Identifying cell-type-specific expression quantitative trait loci (eQTLs) is important to understanding the genetic regulation of gene expressions at the cell-type level and its relevance to complex traits. However, existing eQTL fine-mapping methods are limited in power and accuracy when cell types are analyzed separately. To improve eQTL mapping, we present CASE, a Bayesian framework to perform Cell-type-specific And Shared EQTL fine-mapping that simultaneously analyzes multiple cell types. CASE can effectively capture effect-sharing patterns across cell types while disentangling the confounding effects of linkage disequilibrium (LD). We demonstrate that CASE outperforms the existing single-trait (SuSiE) and multi-trait (mvSuSiE) eQTL methods through comprehensive simulations. When applied to the OneK1K data, CASE identified more genetic regulations of gene expressions, better capturing cell type specificity and functionally enriched and disease-associated eQTLs. The CASE framework for single-cell eQTL fine-mapping can be broadly applied to multi-tissue and multi-trait genetic studies.

genetics↗

bioRxiv2025

Elucidating the Characteristics and Clonal Evolutionary Trajectory of Influenza Neuraminidase Broadly Reactive B Cell

Influenza virus neuraminidase (NA) is receiving increasing attention as a target for universal flu vaccines. Several broad NA inhibition monoclonal antibodies (BImAbs) targeting the highly conserved enzymatic pocket have been previously described. However, the molecular characteristics, clonal evolutionary trajectory, and B cell sources of BImAbs remain poorly understood. Here, using NA-mutant probes, we comprehensively profiled the immune signatures of NA-specific memory B cells (MBCs) from a healthy individual with NA cross-inhibition activity. From the NA-specific MBC repertoires, we identified a series of NA BImAbs with molecular features characterized by long HCDR3 regions with an "xxxDRxxx" motif, which exhibited broad inhibition against diverse influenza NAs. Clonal lineage tracing revealed that these BImAbs followed a clonal evolutionary trajectory encompassing classical MBC (cMBC) and atypical MBC (aMBC). Both cMBC- and aMBC-derived NA BImAbs displayed similar inhibitory activity against the influenza NAs. These findings enhance our understanding of the development of NA BImAbs and provide a foundation for the rational design of NA-based universal flu vaccines.

immunology↗

bioRxiv2025

Bridging Gaps in Antibody Responses and Animal Welfare: Assessing Blood Collection Methods and Vaginal Immunity in Mice Immunized with Intranasal Gonococcal Vaccines

Assessing antibody titers and functional responses is essential for evaluating vaccine efficacy, yet the impact of blood collection methods on these immunological assessments remains unclear. Retro-orbital (RO) blood collection is commonly used but significant complications can occur. Increasingly, investigators have adopted alternative blood collection approaches, such as saphenous vein (SV) sampling to improve laboratory animal welfare. This study compared RO and SV sampling in the development of a Neisseria gonorrhoeae (Ng) vaccine, evaluating Adhesin Complex Protein (ACP) and multiple transferable resistance (Mtr) E protein (MtrE) as antigen candidates. Epitope mapping revealed that ACP and MtrE possess multiple, highly accessible B-cell and T-cell epitope clusters, reinforcing their immunological potential. Following intranasal immunization with rACP, rACP+CpG, and rMtrE+CpG, we assessed the specificity, magnitude, kinetics, and functional quality of immune responses elicited by the immunization regimens. Out of 45 comparisons, only eight significant differences were detected in antibody titers, while the human serum bactericidal assays revealed no differences between RO and SV in antigen-immunized groups. However, antibodies elicited by rACP alone or ACP+CpG in SV samples restored 30.05% and 75.2% of human lysozyme hydrolytic activity compared to 19.3 and 59.9 % in RO, respectively suggesting that SV sampling may be more reliable for assessing functional antibody responses. Beyond its immunological advantages, SV sampling reduces stress, minimizes ocular trauma, and improves animal welfare, making it a viable alternative to RO collection. Given its widespread use in vaccine research, standardizing SV sampling could improve data reliability, ethical compliance, and translational relevance in preclinical studies.

microbiology↗

bioRxiv2025

Elucidation and de novo Reconstitution of Glyceollins Biosynthesis

Glyceollins are phytoalexins produced by soybeans in response to stressors such as pathogen invasion, injury, and environmental challenges. In addition to their antibacterial and antifungal activities, these compounds have attracted significant attention for their potential anticancer, anti-inflammatory, and antioxidant properties. However, their limited accessibility--due to the challenges and high costs associated with organic synthesis or purification from treated soybean seedlings--has hindered further physiological and biochemical studies. Moreover, the incomplete understanding of glyceollin biosynthesis, particularly the final cyclization steps, remains a major barrier to elucidating the physiological functions of glyceollins and achieving sustainable production through synthetic biology. In this study, we identified previously uncharacterized genes encoding two reductases for 7,2,4-trihydroxyisoflavanol (THI) biosynthesis and five P450 enzymes responsible for the final oxidative cyclization in glyceollin I, II, and III biosynthesis, thereby completing the entire glyceollin biosynthetic pathway. By reconstructing the pathway de novo through synthetic biology, we achieved the successful production of glyceollins from simple carbon sources in bakers yeast. This work advances the understanding of glyceollin biosynthesis in soybean, facilitates sustainable production in microbial hosts and offers new opportunities for their application in agriculture and biology.

biochemistry↗

bioRxiv2025

Light-activated tetanus neurotoxin for conditional proteolysis and inducible synaptic inhibition in vivo

The light chain of tetanus neurotoxin (TeNT) is a 52 kD metalloprotease that potently inhibits synaptic transmission by cleaving the endogenous vesicle fusion protein VAMP2. To mitigate the toxicity of TeNT and harness it as a conditional tool for neuroscience, we engineered Light-Activated TeNT (LATeNT) via insertion of the light-sensitive LOV domain into an allosteric site. LATeNT was optimized by directed evolution and shown to have undetectable activity in the dark mammalian brain. Following 30 seconds of weak blue light exposure, however, LATeNT potently inhibited synaptic transmission in multiple brain regions. The effect could be reversed over 24 hours. We used LATeNT to discover an interneuron population in hippocampus that controls anxiety-like behaviors in mouse, and to control the secretion of endogenous insulin from pancreatic beta cells. Synthetic circuits incorporating LATeNT converted drug, Ca2+, or receptor activation into transgene expression or reporter protein secretion. Due to its large dynamic range, rapid kinetics, and highly specific mechanism of action, LATeNT should be a robust tool for conditional proteolysis and spatiotemporal control of synaptic transmission in vivo.

biochemistry↗

bioRxiv2025

Cycling Molecular Assemblies for Selective Cancer Cell Golgi Disruption

The Golgi apparatus is a critical organelle responsible for intracellular trafficking and signaling, orchestrating essential processes such as protein and lipid sorting1-5. Dysregulation of its function has been implicated in various pathologies, including obesity, diabetes, and cancer, highlighting its importance as a potential therapeutic target. Despite this, the development of tools to selectively target the Golgi in specific cell types remain a significant unmet challenge in imaging and drug discovery. Golgi-specific enzyme activities, such as those mediated by protein acyltransferases and thioesterases6, offer an untapped opportunity to develop subcellularly localized therapeutics. Current approaches predominantly rely on direct protein binding but lack the necessary cell selectivity7, underscoring the unmet need for innovative strategies to selectively disrupt Golgi function in cancer cells. Here, we report the development of cycling molecular assemblies (CyMA), a novel class of small peptide derivatives (e.g., dipeptides), which exploit the unique enzymatic environment of the Golgi to establish futile cycles of reversible S-acylation. These assemblies selectively accumulate in cancer cell Golgi, interfering with protein S-acylation cycles and disrupting organelle homeostasis. CyMA impair key Golgi functions, including protein trafficking, glycosylation, and secretion, while demonstrating selective sparing hepatocytes and immune cells such as M1 macrophages. This selective activity represents a paradigm shift, utilizing an enzyme switch and leveraging intracellular environment rather than direct protein binding. Unlike conventional approaches, CyMA reduce tumor growth, drug resistance, and metastasis by pleiotropically disrupting Golgi related functions. By demonstrating the potential of futile cycles as a therapeutic strategy8, this study introduces a generalizable method for targeting organelle-specific enzyme activities. These findings not only underscore the therapeutic potential of CyMA in cancer but also pave the way for future applications in other Golgi-associated diseases.

biochemistry↗

bioRxiv2024

Direct Feature Identification from Raman Spectra and Precise Data-driven Classification of Phytopathogens at Single Conidium-Species Level

Conidia can cause outbreaks of fungal plant diseases, resulting in significant losses in crop yield and economy. Traditional diagnosis of phytopathogenic conidia based on morphology and molecular biology is time-consuming, labor-intensive, and often fails to differentiate fungal conidia. To overcome these challenges, a new classification approach was developed in this study by integrating Raman spectroscopy with data-driven modeling. Eight fungal species were selected and characterized using Raman spectra. Three characteristic Raman wavenumbers at 1003 to 1005 cm-1, 1153 to 1157 cm-1, and 1515 to 1522 cm-1 shared a consistent pattern across species and could be attributed to carotenoids. Clustering of the Raman spectra using principal component analysis (PCA) showed substantial overlap, indicating inaccurate classification of conidia. Three data-driven models, support vector machines (SVMs), decision trees (DTs), and eXtreme Gradient Boosting Forest (XGBoost) were trained with three categories of features (number of peaks, maximum peak, and curve roughness) identified within eight characteristic wavenumber ranges, The optimal SVM, DT, and XGBoost determined by hyperparameter tuning achieved prediction precision of 0.88, 0.88, and 0.96, respectively. PCA-XGBoost trained by feeding principal components of PCA to XGBoost achieved prediction precision of 0.94, suggesting that features extracted from the raw datasets outperformed those extracted with PCA in terms of data-driven classification. This study has demonstrated the great potential of Raman spectroscopy combined with data-driven modeling for classification of phytopathogenic conidia.

microbiology↗

bioRxiv2024

Ethanol alters mechanosensory habituation in C. elegans by way of the BK potassium channel through a novel mechanism

In this research, we investigated how alcohol modulates the simplest form of learning, habituation, in Caenorhabditis elegans. We used our high throughput Multi-Worm Tracker to conduct a large scale study of more than 21,000 wild-type worms to assess the effects of different doses of alcohol on habituation of the well-characterized tap withdrawal response. We found that the effect of alcohol on habituation of the reversal response to repeated mechanosensory stimuli (taps) differed depending on the component of the reversal response assessed. Furthermore, we discovered that alcohol shifted the dominant response to tap from a backward reversal to a brief forward movement. Because the large conductance potassium (BK) channel has been shown to be important for alcohols effects on behaviour in a variety of organisms, including C. elegans, we investigated whether the C. elegans BK channel ortholog, SLO-1, mediated the effects of alcohol on habituation. We tested several different strains of worms with mutations in slo-1 along with wild-type controls; null mutations in slo-1 made animals resistant to alcohol induced changes in learning. However, a mutation in the putative ethanol binding site on SLO-1 did not disrupt ethanols impact on habituation. Finally, by degrading SLO-1 in different parts of the nervous system we found that SLO-1s function in ethanols impact on habituation is likely distributed throughout the neural circuit that responds to tap. Based on these results, our main conclusions are 1) ethanol is not a general facilitator or inhibitor of habituation but rather a complex modulator, 2) SLO-1 is required for ethanols effect on habituation, 3) ethanol is interacting (directly or indirectly) with SLO-1 through a novel unidentified mechanism to influence response plasticity.

neuroscience↗

bioRxiv2024

Chaperonin TRiC bridges radial spokes for folding proteins locally translated in mammalian sperm flagella

As animals evolved from external to internal fertilization, sperm flagella, which once transiently propelled sperm in water to reach nearby eggs, developed to beat for days or even longer within the female reproductive tract. How flagella were remodeled accordingly remains unclear. Unlike externally fertilizing zebrafish and sea urchin, mammalian sperm flagella feature a unique barrel-shaped bridge between axonemal radial spokes RS1 and RS2, though its components and function remain unknown. Here, using cryo-EM and cryo-ET combined with mass spectrometry and immunofluorescence staining, we reveal that this RS1-RS2 barrel (RRB) in mammalian sperm flagella is a group II chaperonin TRiC/CCT. We purified and resolved the unforeseen cryo-EM structure of TRiC from bovine sperm flagella, discovering that flagella TRiC contains the testis-specific CCT6B subunit. Additionally, we resolved the cryo-ET map of mouse sperm flagella RSs, containing the RRB, achieving unprecedented local resolution of 10-14 [A]. The in-situ RRB map displays two extra densities within its chambers in a polarized manner, suggesting the presence of folding substrates and a co-chaperone (likely phosducin-like 2). Notably, mammalian flagella also contain components of translation machineries and actively synthesize proteins. Our results suggest that the RRB TRiC folds locally translated proteins to sustain mammalian flagella, providing new insights into flagellar remodeling in internally fertilizing species. Our findings also shed lights on male fertility and potential treatments for infertility.

biophysics↗

bioRxiv2024

Neutrophil extracellular traps promote tumor chemoresistance to anthracyclines

The microenvironment plays an important role in promoting tumor cell chemoresistance, but the mechanisms responsible for this effect are not clear. Here, using models of multiple myeloma (MM) and solid cancers, we demonstrate a novel mechanism mediated by neutrophils, a major cell population in the bone marrow (BM), that protects cancer cells from chemotherapeutics. We show that in response to tumor-derived soluble factors, BM neutrophils release their DNA in the form of neutrophil extracellular traps (NETs). Cell-free DNA derived from NETs is then taken up by tumor cells via endocytosis and localizes to the cytoplasm. We found that both NETs and cell-free DNA taken up by tumor cells can bind anthracyclines, leading to tumor cell resistance to this class of chemotherapeutic agents. Targeting cell-free DNA with Pulmozyme or blocking NET formation with a PAD4 inhibitor abrogates the chemoprotective effect of neutrophils and restores sensitivity of tumor cells to anthracyclines.

cancer biology↗

bioRxiv2024

DNA nanodevice for analysis of force-activated protein extension and interactions

AbstractForce-induced changes in protein structure and function mediate cellular responses to mechanical stresses. Existing methods to study protein conformation under mechanical force are incompatible with biochemical and structural analysis. Taking advantage of DNA nanotechnology, including the well-defined geometry of DNA origami and programmable mechanics of DNA hairpins, we built a nanodevice to apply controlled forces to proteins. This device was used to study the R1-R2 segment of the talin1 rod domain as a model protein, which comprises two alpha-helical bundles that reversibly unfold under tension to expose binding sites for the cytoskeletal protein vinculin. Electron microscopy confirmed tension-dependent protein extension while biochemical analysis demonstrated enhanced vinculin binding under tension. The device could also be used in pull down assays with cell lysates, which identified filamins as novel tension-dependent talin binders. The DNA nanodevice is thus a valuable addition to the molecular toolbox for studying mechanosensitive proteins.

biophysics↗

bioRxiv2024

OsWHY1/OsTRXz/OsMORFs complex is essential for RNA modification and early chloroplast development in rice

WHIRLY (WHY) proteins are single-stranded DNA/RNA-binding proteins that play multifaceted roles in various plant species. The regulatory mechanisms of WHY proteins in rice remains blank. Here we demonstrate that OsWHY1 in rice is required for early chloroplast development. CRISPR/Cas9-generated oswhy1 knockout lines displayed albino seedling phenotypes, abnormal chloroplast structure and comprised redox balance in leaves. OsWHY1 interacts with multiple plastid proteins, including the thioredoxin OsTRXz and two multiple organellar RNA editing factors (OsMORF8 and OsMORF9) in chloroplasts. Accordingly, several plastid genes dependent on plastid-encoded RNA polymerase (PEP) in the oswhy1 mutants were significantly depressed at both transcript and protein levels. The editing of rps14 transcripts and splicing of rpl2, along with their protein expression, were defective in the oswhy1 mutants. OsWHY1 exhibited RNA-binding activity, specifically binding to rps14 and rpl2 precursor RNAs, which underscores its role as a post-transcriptional regulator essential for normal protein synthesis in chloroplasts. Loss-of- function mutants of either OsWHY1 or OsMORF9 and OsTRXz displayed albino phenotypes, disrupted H2O2 homeostasis, and defective RNA processing in rps14 and rpl2, suggesting the OsWHY1-OsTRXz-OsMORFs regulatory module is vital for maintaining chloroplast stability and integrity through its RNA-binding activity and its role in recruiting OsTRXz and OsMORFs to ensure proper RNA modification. One sentence summaryOsWHY1 is integral to chloroplast development in rice through its RNA-binding activity and its role in recruiting OsTRXz and OsMORFs to ensure proper RNA modification.

plant biology↗

bioRxiv2024

In situ deposition of nanobodies by an engineered commensal microbe promotes survival in a mouse model of enterohemorrhagic E. coli

Engineered smart microbes that deliver therapeutic payloads are emerging as treatment modalities, particularly for diseases with links to the gastrointestinal tract. Enterohemorrhagic E coli (EHEC) is a causative agent of potentially lethal hemolytic uremic syndrome. Given concerns that antibiotic treatment increases EHEC production of Shiga toxin (Stx), which is responsible for systemic disease, novel remedies are needed. EHEC encodes a type III secretion system (T3SS) that injects Tir into enterocytes. Tir inserts into the host cell membrane, exposing an extracellular domain that subsequently binds intimin, one of its outer membrane proteins, triggering the formation of attaching and effacing (A/E) lesions that promote EHEC mucosal colonization. Citrobacter rodentium (Cr), a natural A/E mouse pathogen, similarly requires Tir and intimin for its pathogenesis. Mice infected with Cr({Phi}Stx2dact), a variant lysogenized with an EHEC-derived phage that produces Stx2dact, develop intestinal A/E lesions and toxin-dependent disease. Stx2a is more closely associated with human disease. By developing an efficient approach to seamlessly modify the C. rodentium genome, we generated Cr_Tir-MEHEC({Phi}Stx2a), a variant that expresses Stx2a and the EHEC extracellular Tir domain. We found that mouse pre-colonization with HS-PROT3EcT-TD4, a human commensal E. coli strain (E. coli HS) engineered to efficiently secrete- an anti-EHEC Tir nanobody, delayed bacterial colonization and improved survival after challenge with Cr_Tir-MEHEC({Phi}Stx2a). This study provides the first evidence to support the efficacy of engineered commensal E. coli to intestinally deliver therapeutic payloads that block essential enteric pathogen virulence determinants, a strategy that may serve as an antibiotic-independent antibacterial therapeutic modality. Significance StatementEngineered smart microbes that secrete therapeutics are emerging as treatment modalities, particularly for gut-based diseases. With the growing threat of multidrug-resistant infection, non-antibiotic treatments are urgently needed. The gastrointestinal pathogen enterohemorrhagic E coli (EHEC) can cause the potentially lethal hemolytic uremic syndrome, a toxin-driven disease. Given concerns that antibiotics increase toxin release, treatment is largely limited to supportive care. Here, we show that pre-treatment with a commensal E. coli (HS-PROT3EcT) engineered to secrete an antibody that blocks an essential EHEC virulence factor delays the establishment of an EHEC-like infection in mice. This study strongly suggests that smart microbes that deliver payloads that block colonization factors of gut pathogens can be developed as critically needed alternatives to antibiotics for fighting bacterial infections.

microbiology↗

bioRxiv2024

ChromBERT: Uncovering Chromatin State Motifs in the Human Genome Using a BERT-based Approach

Chromatin states, which are defined by specific combinations of histone post-translational modifications, are fundamental to gene regulation and cellular identity. Despite their importance, comprehensive patterns within chromatin state sequences, which could provide insights into key biological functions, remain largely unexplored. In this study, we introduce ChromBERT, a BERT-based model specifically designed to detect distinct chromatin state patterns as "motifs." We pre-trained ChromBERT on 15-state chromatin annotations from 127 human cell and tissue types from the ROADMAP consortium. This pre-trained model can be fine-tuned for various downstream tasks, and obtained high-attention chromatin state patterns are extracted as motifs. To account for the variable-length nature of chromatin state motifs, ChromBERT uses Dynamic Time Warping to cluster similar motifs and identify meaningful representative patterns. In this study, we evaluated the performance of the model on several tasks, including binary and quantitative gene expression prediction, cell type classification, and three-dimensional genome feature classification. Our analyses yielded biologically grounded results and revealed the associated chromatin state motifs. This workflow facilitates the discovery of specific chromatin state patterns across different biological contexts and offers a new framework for exploring the dynamics of epigenomic states.

bioinformatics↗

bioRxiv2024

Muskelin acts as a substrate receptor of the highly regulated Drosophila CTLH E3 ligase during the maternal-to-zygotic transition

The maternal-to-zygotic transition (MZT) is a conserved developmental process where the maternally-derived protein and mRNA cache is replaced with newly made zygotic gene products. We have previously shown that in Drosophila the deposited RNA-binding proteins ME31B, Cup, and Trailer Hitch (TRAL) are ubiquitylated by the CTLH E3 ligase and cleared. However, the organization and regulation of the CTLH complex remain poorly understood in flies. In particular, Drosophila lacks an identifiable substrate adaptor, and the mechanisms restricting degradation of ME31B and its cofactors to the MZT are unknown. Here, we show that the developmental specificity of the CTLH complex is mediated by multi-pronged regulation, including transcriptional control by the transcription factor OVO and autoinhibition of the E3 ligase. One major regulatory target is the subunit Muskelin, which we demonstrate acts as a substrate adaptor for the Drosophila CTLH complex. Although conserved, Muskelin has structural roles in other species, suggesting a surprising functional plasticity. Finally, we find that Muskelin has few targets beyond the three known RNA binding proteins, showing exquisite target specificity. Thus, multiple levels of integrated regulation restrict the activity of the embryonic CTLH complex to early embryogenesis, seemingly with the goal of regulating three important RNA binding proteins.

developmental biology↗

bioRxiv2024

Functional imaging of whole mouse embryonic development in utero

Despite significant advances in structural and genetic studies, investigations of early embryonic functions, such as brain activity, have long been constrained by technical challenges. Functional ultrasound (fUS) has emerged as a breakthrough modality, enabling real-time monitoring of brain activity with exceptional spatial and temporal resolution and offering unprecedented opportunities for studying functional embryonic development. In this study, we used fUS to monitor whole-embryo activity in mice from embryonic days E8.5 to E18.5, revealing patterns of neural activity throughout embryogenesis. This approach provides new opportunities to explore brain development dynamically as it unfolds. Moreover, we observed embryo responses to external stimuli, including sound, in both mice and cynomolgus macaques, offering insights into early sensory processing and neural maturation. In summary, our study establishes fUS as a powerful tool for studying embryonic brain functional development, with significant implications for scientific research, especially in non- human primate models, and clinical applications.

developmental biology↗

bioRxiv2024

A DNA-based molecular clamp for probing protein interactions and structure under force.

Cellular mechanotransduction, a process central to cell biology, embryogenesis, adult physiology and multiple diseases, is thought to be mediated by force-driven changes in protein conformation that control protein function. However, methods to study proteins under defined mechanical loads on a biochemical scale are lacking. We report the development of a DNA based device in which the transition between single-stranded and double-stranded DNA applies tension to an attached protein. Using a fragment of the talin rod domain as a test case, negative-stain electron microscopy reveals programmable extension while pull down assays show tension-induced binding to two ligands, ARPC5L and vinculin, known to bind to cryptic sites inside the talin structure. These results demonstrate the utility of the DNA clamp for biochemical studies and potential structural analysis.

bioengineering↗

bioRxiv2024

Predicting Primary Graft Dysfunction in Lung Transplantation: Machine Learning-Guided Biomarker Discovery

BACKGROUNDThere is an urgent need to better understand the pathophysiology of primary graft dysfunction (PGD) so that point-of-care methods can be developed to predict those at risk. Here we utilize a multiplex multivariable approach to define cytokine, chemokines, and growth factors in patient-matched biospecimens from multiple biological sites to identify factors predictive of PGD. METHODSBiospecimens were collected from patients undergoing bilateral LTx from three distinct sites: donor lung perfusate, post-transplant bronchoalveolar lavage (BAL) fluid (2h), and plasma (2h and 24h). A 71-multiplex panel was performed on each biospecimen. Cross-validated logistic regression (LR) and random forest (RF) machine learning models were used to determine whether analytes in each site or from combination of sites, with or without clinical data, could discriminate between PGD grade 0 (n = 9) and 3 (n = 8). RESULTSUsing optimal AUROC, BAL fluid at 2h was the most predictive of PGD (LR, 0.825; RF, 0.919), followed by multi-timepoint plasma (LR, 0.841; RF, 0.653), then perfusate (LR, 0.565; RF, 0.448). Combined clinical, BAL, and plasma data yielded strongest performance (LR, 1.000; RF, 1.000). Using a LASSO of the predictors obtained using LR, we selected IL-1RA, BCA-1, and Fractalkine, as most predictive of severe PGD. CONCLUSIONSBAL samples collected 2h post-transplant were the strongest predictors of severe PGD. Our machine learning approach not only identified novel cytokines not previously associated with PGD, but identified analytes that could be used as a point-of-care cytokine panel aimed at identifying those at risk for developing severe PGD.

immunology↗