Biology subjects

Fu, Z.

Publications and source records attributed to Fu, Z..

At least 145 records · Page 8Linked to original sources

bioRxiv2021

Two interacting transcriptional coactivators cooperatively control plant immune responses

The phytohormone salicylic acid (SA) plays a pivotal role in plant defense against biotrophic and hemibiotrophic pathogens. Genetic studies have identified NPR1 and EDS1 as two central hubs in plant local and systemic immunity. However, it is unclear how NPR1 orchestrates gene regulation and whether EDS1 directly participates in transcriptional reprogramming. Here we show that NPR1 and EDS1 synergistically activate Pathogenesis-Related (PR) genes and plant defenses by forming a protein complex and co-opting with Mediator. In particular, we discover that EDS1 functions as an autonomous transcriptional coactivator with intrinsic transactivation domains and physically interacts with the CDK8 subunit of Mediator. Upon SA induction, EDS1 is directly recruited by NPR1 onto the PR1 promoter via physical NPR1-EDS1 interactions, thereby potentiating PR1 activation. We further demonstrate that EDS1 stabilizes NPR1 protein and NPR1 transcriptionally upregulates EDS1 in plant-pathogen interactions. Our results reveal an elegant interplay of key coactivators with Mediator and elucidate novel molecular mechanisms for activating transcription during immune responses.

plant biology↗

bioRxiv2021

Individualized Spatial Network Predictions Using Siamese Convolutional Neural Networks: A Resting-State fMRI Study of over 11,000 Unaffected Individuals

Individuals can be characterized in a population according to their brain measurements and activity, given the inter-subject variability in brain anatomy, structure-function relationships, or life experience. Many neuroimaging studies have demonstrated the potential of functional network connectivity patterns estimated from resting functional magnetic resonance imaging (fMRI) to discriminate groups and predict information about individual subjects. However, the predictive signal present in the spatial heterogeneity of brain connectivity networks is yet to be extensively studied. In this study, we investigate, for the first time, the use of pairwise-relationships between resting-state independent spatial maps to characterize individuals. To do this, we develop a deep Siamese framework comprising three-dimensional convolution neural networks for contrastive learning based on individual-level spatial maps estimated via a fully automated fMRI independent component analysis approach. The proposed framework evaluates whether pairs of spatial networks (e.g., visual network and auditory network) are capable of subject identification and assesses the spatial variability in different network pairs predictive power in an extensive whole-brain analysis. Our analysis on nearly 12,000 unaffected individuals from the UK Biobank study demonstrates that the proposed approach can discriminate subjects with an accuracy of up to 88% for a single network pair on the test set (best model, after several runs), and 82% average accuracy at the subcortical domain level, notably the highest average domain level accuracy attained. Further investigation of our networks learned features revealed a higher spatial variability in predictive accuracy among younger brains and significantly higher discriminative power among males. In sum, the relationship among spatial networks appears to be both informative and discriminative of individuals and should be studied further as putative brain-based biomarkers.

neuroscience↗

bioRxiv2021

GABAergic neuronal IL-4R mediates T cell effect on memory

Mechanisms governing how immune cells and their derived molecules impact homeostatic brain function are still poorly understood. Here, we elucidate neuronal mechanisms underlying effects of T cells on synaptic function and episodic memory. Depletion of CD4 T cells led to memory deficits and impaired long term potentiation. Severe combined immune-deficient (SCID) mice exhibited amnesia, which was reversible by repopulation with T cells from wild-type but not from IL-4-knockout mice. This rescue was mediated via IL-4 receptors (IL-4R) expressed on neurons. Exploration of snRNAseq of neurons participating in memory processing provided insights into synaptic organization and plasticity-associated pathways and genes regulated by immune cells and molecules. IL-4R knockout in inhibitory neurons impaired contextual fear memory, suggesting participation of an IL-4-associated switch in regulating synaptic function and promoting contextual fear memory. These findings provide insights into neuroimmune interactions at the transcriptional and functional levels in neurons.

neuroscience↗

bioRxiv2021

An updated assembly strategy helps parsing the cryptic mitochondrial genome evolution in plants

Although plant mitogenomes are small in size, their variations are no less than any other complex genomes. They are under rapid structure and size changes. These characters make the assembly a great challenge. This caused two intertwined problems, a slow growth of known mitogenomes and a poor knowledge of their evolution. In many species, mitogenome becomes the last genome that undeciphered. To have a better understanding of these two questions, we developed a strategy using short sequencing reads and combining current tools and manual steps to get high quality mitogenomes. This strategy allowed us to assembled 23 complete mitogenomes from 5 families in Fagales. Our large-scale comparative genomic analyses indicated the composition of mitogenomes is very mosaic that "horizontal transfers" can be from almost all taxa in seed plants. The largest mitogenome contains more homologous DNA with other Fagales, rather than unique sequences. Besides of real HGTs, sometimes mitovirus, nuclear insertions and other third-part DNA could also produce HGT-like sequences, accounting partially for the unusual evolutionary trajectories, including the cryptic size expansion in Carpinus. Mitochondrial plasmid was also found. Its lower GC content indicates that it may be only an interphase of a foreign DNA before accepting by the main chromosome. Our methods and results provide new insights into the assembly and mechanisms of mitogenome evolution.

evolutionary biology↗

bioRxiv2021

Molecular mechanism of ACAD9 in mitochondrial respiratory complex 1 assembly

ACAD9 belongs to the acyl-CoA dehydrogenase family, which catalyzes the -{beta} dehydrogenation of fatty acyl-CoA thioesters. Thus, it is involved in fatty acid {beta}-oxidation (FAO). However, it is now known that the primary function of ACAD9 is as an essential chaperone for mitochondrial respiratory complex 1 assembly. ACAD9 interacts with ECSIT and NDUFAF1, forming the mitochondrial complex 1 assembly (MCIA) complex. Although the role of MCIA in the complex 1 assembly pathway is well studied, little is known about the molecular mechanism of the interactions among these three assembly factors. Our current studies reveal that when ECSIT interacts with ACAD9, the flavoenzyme loses the FAD cofactor and consequently loses its FAO activity, demonstrating that the two roles of ACAD9 are not compatible. ACAD9 binds to the carboxy-terminal half (C-ECSIT), and NDUFAF1 binds to the amino-terminal half of ECSIT. Although the binary complex of ACAD9 with ECSIT or with C-ECSIT is unstable and aggregates easily, the ternary complex of ACAD9-ECSIT-NDUFAF1 (i.e., the MCIA complex) is soluble and extremely stable. Molecular modeling and SAXS studies of the MCIA complex identified the possible interaction sites between the three assembly factors and binding sites for other assembly factors, including complex 1 subunits. Furthermore, we have mapped over 40 currently known pathogenic mutation sites onto the homology-modeled ACAD9 structure, giving us the structural basis for their involvement in diseases that result from complex 1 deficiency.

biochemistry↗

bioRxiv2021

Multi-Spatial Scale Dynamic Interactions between Functional Sources Reveal Sex-Specific Changes in Schizophrenia

We introduce an extension of independent component analysis (ICA), called multiscale ICA (msICA), and design an approach to capture dynamic functional source interactions within and between multiple spatial scales. msICA estimates functional sources at multiple spatial scales without imposing direct constraints on the size of functional sources, overcomes the limitation of using fixed anatomical locations, and eliminates the need for model-order selection in ICA analysis. We leveraged this approach to study sex-specific and -common connectivity patterns in schizophrenia. Results show dynamic reconfiguration and interaction within and between multi-spatial scales. Sex-specific differences occur (1) within the subcortical domain, (2) between the somatomotor and cerebellum domains, and (3) between the temporal domain and several others, including the subcortical, visual, and default mode domains. Most of the sex-specific differences belong to between-spatial scale functional interactions and are associated with a dynamic state with strong functional interactions between the visual, somatomotor, and temporal domains and their anticorrelation patterns with the rest of the brain. We observed significant correlations between multi-spatial scale functional interactions and symptom scores, highlighting the importance of multiscale analyses to identify potential biomarkers for schizophrenia. As such, we recommend such analyses as an important option for future functional connectivity studies.

neuroscience↗

bioRxiv2021

Aberrant Dynamic Functional Connectivity of Default Mode Network in Schizophrenia and Links to Symptom Severity

BackgroundSchizophrenia affects around 1% of the global population. Functional connectivity extracted from resting-state functional magnetic resonance imaging (rs-fMRI) has previously been used to study schizophrenia and has great potential to provide novel insights into the disorder. Some studies have shown abnormal functional connectivity in the default mode network of individuals with schizophrenia, and more recent studies have shown abnormal dynamic functional connectivity (dFC) in individuals with schizophrenia. However, DMN dFC and the link between abnormal DMN dFC and symptom severity have not been well characterized. MethodResting-state fMRI data from subjects with schizophrenia (SZ) and healthy controls (HC) across two datasets were analyzed independently. We captured seven maximally independent subnodes in the DMN by applying group independent component analysis and estimated dFC between subnode time courses using a sliding window approach. A clustering method separated the dFCs into five reoccurring brain states. A feature selection method modeled the difference between SZs and HCs using the state-specific FC features. Finally, we used the transition probability of a hidden Markov model to characterize the link between symptom severity and dFC in SZ subjects. ResultsWe found decreases in the connectivity of the anterior cingulate cortex (ACC) and increases in the connectivity between the precuneus (PCu) and the posterior cingulate cortex (PCC) (i.e., PCu/PCC) of SZ subjects. In SZ, the transition probability from a state with weaker PCu/PCC and stronger ACC connectivity to a state with stronger PCu/PCC and weaker ACC connectivity increased with symptom severity. ConclusionsTo our knowledge, this was the first study to investigate DMN dFC and its link to schizophrenia symptom severity. We identified reproducible neural states in a data-driven manner and demonstrated that the strength of connectivity within those states differed between SZs and HCs. Additionally, we identified a relationship between SZ symptom severity and the dynamics of DMN functional connectivity. We validated our results across two datasets. These results support the potential of dFC for use as a biomarker of schizophrenia and shed new light upon the relationship between schizophrenia and DMN dynamics.

neuroscience↗

bioRxiv2021

Disrupted dynamic functional network connectivity among cognitive control networks in the progression of Alzheimer's disease

BackgroundAlzheimers disease (AD) is the most common age-related dementia that promotes a decline in memory, thinking, and social skills. The initial stages of dementia can be associated with mild symptoms, and symptom progression to a more severe state is heterogeneous across patients. Recent work has demonstrated the potential for functional network mapping to assist in the prediction of symptomatic progression. However, this work has primarily used static functional connectivity (sFC) from rs-fMRI. Recently, dynamic functional connectivity (dFC) has been recognized as a powerful advance in functional connectivity methodology to differentiate brain network dynamics between healthy and diseased populations. MethodsGroup independent component analysis was applied to extract 17 components within the cognitive control network (CCN) from 1385 individuals across varying stages of AD symptomology. We estimated dFC among 17 components within the CCN, followed by clustering the dFCs into 3 recurring brain states and then estimated a hidden Markov model and the occupancy rate for each subject. Finally, we investigated the link between CCN dFC connectivity features with AD progression. ResultsProgression of AD symptoms were associated with increases in connectivity within the middle frontal gyrus. Also, the AD with mild and severer symptoms showed less connectivity within the inferior parietal lobule and between this region with the rest of CCN. Finally, comparing with mild dementia, we found that the normal brain spends significantly more time in a state with lower within middle frontal gyrus connectivity and higher connectivity between the hippocampus and the rest of CCN, highlighting the importance of assessing the dynamics of brain connectivity in this disease. ConclusionOur results suggest that AD progress not only alters the CCN connectivity strength but also changes the temporal properties in this brain network. This suggests the temporal and spatial pattern of CCN as a biomarker that differentiates different stages of AD. Impact StatementBy assuming that functional connectivity is static over time, many of previous studies have ignored the brain dynamic in Alzheimers disease progression. Here, a longitudinal resting-state functional magnetic resonance imaging data are used to explore the temporal changes of functional connectivity in the cognitive control network in Alzheimers disease progression. The result of this study would increase our understanding about the underlying mechanisms of Alzheimers Disease and help in finding future treatment of this neurological disorder.

neuroscience↗

bioRxiv2020

Building Models of Functional Interactions Among Brain Domains that Encode Varying Information Complexity: A Schizophrenia Case Study

Revealing associations among various structural and functional patterns of the brain can yield highly informative results about the healthy and disordered brain. Studies using neuroimaging data have more recently begun to utilize the information within as well as across various functional and anatomical domains (i.e., groups of brain networks). However, most whole-brain approaches assume similar complexity of interactions throughout the brain. Here we investigate the hypothesis that interactions between brain networks capture varying amounts of complexity, and that we can better capture this information by varying the complexity of the model subspace structure based on available training data. To do this, we employ a Bayesian optimization-based framework known as the Tree Parzen Estimator (TPE) to identify, exploit and analyze patterns of variation in the information encoded by temporal information extracted from functional magnetic resonance imaging (fMRI) subdomains of the brain. Using a repeated cross-validation procedure on a schizophrenia classification task, we demonstrate evidence that interactions between specific functional subdomains are better characterized by more sophisticated model architectures compared to less complicated ones required by the others for optimally contributing towards classification and understanding the brains functional interactions. We show that functional subdomains known to be involved in schizophrenia require more complex architectures to optimally unravel discriminatory information about the disorder. Our study points to the need for adaptive, hierarchical learning frameworks that cater differently to the features from different subdomains, not only for a better prediction but also for enabling the identification of features predicting the outcome of interest.

neuroscience↗

bioRxiv2020

Target lipidomics reveals associations between serum sphingolipids and insulin sensitivity by the glucose clamp

BackgroundThis study aimed to systematically investigate the associations between serum sphingolipids and insulin sensitivity as well as insulin secretion. This study also aimed to reveal potential predictors for insulin sensitivity or give perceptive insight into disease processes. MethodsWe conducted a lipidomics evaluation of molecularly distinct SPs in the serum of 86 consecutive Chinese adults with or without obesity and diabetes using electrospray ionization mass spectrometry coupled with liquid chromatography. The GIR30 was measured under steady conditions to assess insulin sensitivity by the gold standard hyperinsulinemic-euglycemic clamp. We created the ROC curves to detect the serum SMs diagnostic value and establish the diagnosis of insulin sensitivity. ResultsDifferential correlation network analysis illustrated correlations amongst lipids, insulin sensitivity, insulin secretion and other clinical indexes. Total and subspecies of serum SMs and globotriaosylceramides (Gb3s) were positively related to GIR30, free FAs (FFA 16:1, FFA20:4), some long chain GM3 and complex ceramide GluCers showed strong negative correlations with GIR30. Notably, ROC curves showed that SM/Cer and SM d18:0/26:0 may be good serum lipid predictors of diagnostic indicators of insulin sensitivity close to conventional clinical indexes such as 1/HOMA-IR (all areas under the curve >0.80) based on GIR30 as standard diagnostic criteria. ConclusionsThese results provide novel associations between serum sphingolipid between insulin sensitivity measured by the hyperinsulinemic-euglycemic clamp. We further identify two specific SPs that may represent prognostic biomarkers for insulin sensitivity.

physiology↗

bioRxiv2020

Mapping Isoform Abundance and Interactome of the Endogenous TMPRSS2-ERG Fusion Protein with Orthogonal Immunoprecipitation-Mass Spectrometry Assays

TMPRSS2-ERG gene fusion, a molecular alteration driving nearly a half of prostate cancer cases, has been intensively characterized at the transcript level, while limited studies explored the molecular identity and function of the endogenous fusion at the protein level. Here, we developed and applied immunoprecipitation-mass spectrometry (IP-MS) assays for the measurement of a low-abundance T1E4 TMPRSS2-ERG fusion protein, its isoforms and its interactome in VCaP prostate cancer cells. IP-MS assays quantified total ERG ([~]27,000 copies/cell) and its four unique isoforms, and revealed that the T1E4-ERG isoform accounts for 71% of the total ERG protein in VCaP cells. For the first time, the N-terminal peptide (methionine-truncated and N-acetylated TASSSSDYGQTSK) unique for the T1/E4 fusion was identified and quantified. IP-MS with the C-terminal antibodies identified 29 proteins in the ERG interactome, including SWI/SNF chromatin remodeling complex subunits and numerous transcriptional co-regulators. Our data also suggested that TMPRSS2-ERG protein-protein interactions were exerted through at least two different regions. Knowledge on the distinct TMPRSS2-ERG protein isoforms and interactomes may facilitate development of more accurate diagnostics and targeted therapeutics of prostate cancer.

molecular biology↗

bioRxiv2020

Genome assembly and transcriptome analysis provide insights into the anti-schistosome mechanism of Microtus fortis

Microtus fortis (M. fortis) so far is the only mammal host that exhibits intrinsic resistance against Schistosoma japonicum infection. However, the underlying molecular mechanisms of this intrinsic resistance are not yet known. Here we performed the first de novo genome assembly of M. fortis, comprehensive gene annotation and evolution analysis. Furthermore, we compared the recovery rate of schistosome, pathological change and liver transcriptome between non-permissive host M. fortis and susceptible host mouse at different time points after Schistosome infection. We reveal that Immune response of M. fortis and mouse is different in time and type. M. fortis activates immune and inflammatory responses on the 10th days post infection, involving in multiple pathways, such as leukocyte extravasation, antibody activation (especially IgG3), Fc-gamma receptor mediated phagocytosis, and interferon signaling cascade. The strong immune responses of M. fortis in early stages of infection play important roles in preventing the development of schistosome. On the contrary, intense immune response occurred in mouse in late stages of infection (28~42 days post infection), and cannot eliminate schistosome. Infected mouse suffers severe pathological injury and continuous decrease of important functions such as cell cycle and lipid metabolism. Our findings offer new insights to the intrinsic resistance mechanism of M. fortis against schistosome infection. The genome sequence also provides bases for future studies of other important traits in M. fortis.

bioinformatics↗

bioRxiv2020

Intracellular accumulation and secretion of hydrophobin-enriched vesicles aid the rapid sporulation of molds

Fungi can rapidly produce large amounts of spores suitable for aerial dispersal. The hydrophobicity of spores is provided by the unique amphiphilic and superior surface-active proteins - hydrophobins (HFBs) - that self-assemble at hydrophobic/hydrophilic interfaces and thus change surface properties. Using the HFB-enriched mold Trichoderma and the HFB-free yeast Pichia pastoris, we revealed a distinctive HFB secretory pathway that includes an intracellular accumulation of HFBs in lipid bodies (LBs) that can internalize in vacuoles. The resulting vacuolar multicisternal structures (VMS) are stabilized by HFB layers that line up on their surfaces. These HFB-enriched VMSs can move to the periplasm for secretion or become fused in large tonoplast-like organelles. The latter contributes to the maintenance of turgor pressure required for the erection of sporogenic structures and rapid HFB secretion by squeezing out periplasmic VMSs through the cell wall. Thus, HFBs are essential accessory proteins for the development of aerial hyphae and colony architecture.

microbiology↗

bioRxiv2020

Modular and state-relevant connectivity in high-frequency resting-state BOLD fMRI data: An independent component analysis

Resting-state fMRI (rs-fMRI) data are typically filtered at different frequency bins between 0.008[~]0.2 Hz (varies across the literature) prior to analysis to mitigate nuisance variables (e.g., drift, motion, cardiac, and respiratory) and maximize the sensitivity to neuronal-mediated BOLD signal. However, multiple lines of evidence suggest meaningful BOLD signal may also be parsed at higher frequencies. To test this notion, a functional network connectivity (FNC) analysis based on a spatially informed independent component analysis (ICA) was performed at seven different bandpass frequency bins to examine FNC matrices across spectra. Further, eyes open (EO) vs. eyes closed (EC) resting-state acquisitions from the same participants were compared across frequency bins to examine if EO vs. EC FNC matrices and randomness estimations of FNC matrices are distinguishable at different frequencies. Results show that FNCs in higher-frequency bins display modular FNC similar to the lowest frequency bin, while r-to-z FNC and FNC-based measures indicating matrix non-randomness were highest in the 0.31-0.46 Hz range relative to all frequency bins above and below this range. As such, the FNC within this range appears to be the most temporally correlated, but the mechanisms facilitating this coherence require further analyses. Compared to EO, EC displayed greater FNC (involved in visual, cognitive control, somatomotor, and auditory domains) and randomness values at lower frequency bins, but this phenomenon flipped (EO > EC) at frequency bins greater than 0.46 Hz, particularly within visual regions. While the effect sizes range from small to large specific to frequency range and resting state (EO vs. EC), with little influence from common artifacts. These differences indicate that unique information can be derived from FNC between BOLD signals at different frequencies relative to a given restingstate acquisition and support the hypothesis meaningful BOLD signal is present at higher frequency ranges.

neuroscience↗

bioRxiv2020

Structural basis for the inhibition of the papain-like protease of SARS-CoV-2 by small molecules

SARS-CoV-2 is the pathogen responsible for the COVID-19 pandemic. The SARS-CoV-2 papain-like cysteine protease has been implicated in virus maturation, dysregulation of host inflammation and antiviral immune responses. We showed that PLpro preferably cleaves the K48-ubiquitin linkage while also being capable of cleaving ISG15 modification. The multiple functions of PLpro render it a promising drug target. Therefore, we screened an FDA-approved drug library and also examined available inhibitors against PLpro. Inhibitor GRL0617 showed a promising IC50 of 2.1 M. The co-crystal structure of SARS-CoV-2 PLpro-C111S in complex with GRL0617 suggests that GRL0617 is a non-covalent inhibitor. NMR data indicate that GRL0617 blocks the binding of ISG15 to PLpro. The antiviral activity of GRL0617 reveal that PLpro is a promising drug target for therapeutically treating COVID-19. One Sentence SummaryCo-crystal structure of PLpro in complex with GRL0617 reveals the druggability of PLpro for SARS-CoV-2 treatment.

biochemistry↗

bioRxiv2020

Phosphosite T674A mutation in kinesin family member 3A fails to reproduce tissue and ciliary defects characteristic of CILK1 loss of function

Background Kinesin family member 3A (KIF3A) is a molecular motor protein in the heterotrimeric kinesin-2 complex that drives anterograde intraflagellar transport. This process plays a pivotal role in both biogenesis and maintenance of the primary cilium that supports tissue development. Ciliogenesis associated kinase 1 (CILK1) phosphorylates human KIF3A at Thr672. CILK1 loss of function causes ciliopathies that manifest profound and multiplex developmental defects, including polydactyly, shortened and hypoplastic bones and alveoli airspace deficiency, leading to perinatal lethality. Prior studies have raised the hypothesis that CILK1 phosphorylation of KIF3A is critical for its regulation of organ development.Results We produced a mouse model with phosphorylation site Thr674 in mouse Kif3a mutated to Ala. Kif3a T674A homozygotes are viable and exhibit no skeletal abnormalities, and only mildly reduced airspace in alveoli. Mouse embryonic fibroblasts carrying Kif3a T674A mutation show a normal rate of ciliation and a moderate increase in cilia length.Conclusion These results indicate that eliminating Kif3a Thr674 phosphorylation by CILK1 is insufficient to reproduce the severe developmental defects in ciliopathies caused by Cilk1 loss of function. This suggests KIF3A phosphorylation by CILK1 is not essential for tissue development and other substrates are involved in Cilk1 ciliopathies.Competing Interest StatementThe authors have declared no competing interest.View Full Text

developmental biology↗

bioRxiv2020

Hype versus hope: Deep learning encodes more predictive and robust brain imaging representations than standard machine learning

Previous successes of deep learning (DL) approaches on several complex tasks have hugely inflated expectations of their power to learn subtle properties of complex brain imaging data, and scale to large datasets. Perhaps as a reaction to this inflation, recent critical commentaries unfavorably compare DL with standard machine learning (SML) approaches for the analysis of brain imaging data. Yet, their conclusions are based on pre-engineered features which deprives DL of its main advantage: representation learning. Here we evaluate this and show the importance of representation learning for DL performance on brain imaging data. We report our findings from a large-scale systematic comparison of SML approaches versus DL profiled in a ten-way age and gender-based classification task on 12,314 structural MRI images. Results show that DL methods, if implemented and trained following the prevalent DL practices, have the potential to substantially improve compared to SML approaches. We also show that DL approaches scale particularly well presenting a lower asymptotic complexity in relative computational time, despite being more complex. Our analysis reveals that the performance improvement saturates as the training sample size grows, but shows significantly higher performance throughout. We also show evidence that the superior performance of DL is primarily due to the excellent representation learning capabilities and that SML methods can perform equally well when operating on representations produced by the trained DL models. Finally, we demonstrate that DL embeddings span a comprehensible projection spectrum and that DL consistently localizes discriminative brain biomarkers, providing an example of the robustness of prediction relevance estimates. Our findings highlight the presence of non-linearities in brain imaging data that DL frameworks can exploit to generate superior predictive representations for characterizing the human brain, even with currently available data sizes.

neuroscience↗

bioRxiv2020

Computational target fishing by mining transcriptional data using a novel Siamese spectral-based graph convolutional network

Computational target fishing aims to investigate the mechanism of action or the side effects of bioactive small molecules. Unfortunately, conventional ligand-based computational methods only explore a confined chemical space, and structure-based methods are limited by the availability of crystal structures. Moreover, these methods cannot describe cellular context-dependent effects and are thus not useful for exploring the targets of drugs in specific cells. To address these challenges, we propose a novel Siamese spectral-based graph convolutional network (SSGCN) model for inferring the protein targets of chemical compounds from gene transcriptional profiles. Although the gene signature of a compound perturbation only provides indirect clues of the interacting targets, the SSGCN model was successfully trained to learn from known compound-target pairs by uncovering the hidden correlations between compound perturbation profiles and gene knockdown profiles. Using a benchmark set, the model achieved impressive target inference results compared with previous methods such as Connectivity Map and ProTINA. More importantly, the powerful generalization ability of the model observed with the external LINCS phase II dataset suggests that the model is an efficient target fishing or repositioning tool for bioactive compounds.

bioinformatics↗