Biology subjects

Fu, Z.

Publications and source records attributed to Fu, Z..

At least 109 records · Page 6Linked to original sources

bioRxiv2023

Replication and Refinement of Brain Age Model for adolescent development.

The discrepancy between chronological age and estimated brain age, known as the brain age gap, may serve as a biomarker to reveal brain development and neuropsychiatric problems. This has motivated many studies focusing on the accurate estimation of brain age using different features and models, of which the generalizability is yet to be tested. Our recent study has demonstrated that conventional machine learning models can achieve high accuracy on brain age prediction during development using only a small set of selected features from multimodal brain imaging data. In the current study, we tested the replicability of various brain age models on the Adolescent Brain Cognitive Development (ABCD) cohort. We proposed a new refined model to improve the robustness of brain age prediction. The direct replication test for existing brain age models derived from the age range of 8-22 years onto the ABCD participants at baseline (9 to 10 years old) and year-two follow-up (11 to 12 years old) indicate that pre-trained models could capture the overall mean age failed precisely estimating brain age variation within a narrow range. The refined model, which combined broad prediction of the pre-trained model and granular information with the narrow age range, achieved the best performance with a mean absolute error of 0.49 and 0.48 years on the baseline and year-two data, respectively. The brain age gap yielded by the refined model showed significant associations with the participants information processing speed and verbal comprehension ability on baseline data.

neuroscience↗

bioRxiv2023

Spatial Dynamic Subspaces Encode Sex-Specific Schizophrenia Disruptions in Transient Network Overlap and its Links to Genetic Risk

BackgroundRecent advances in resting-state fMRI allow us to study spatial dynamics, the phenomenon of brain networks spatially evolving over time. However, most dynamic studies still use subject-specific, spatially-static nodes. As recent studies have demonstrated, incorporating time-resolved spatial properties is crucial for precise functional connectivity estimation and gaining unique insights into brain function. Nevertheless, estimating time-resolved networks poses challenges due to the low signal-to-noise ratio, limited information in short time segments, and uncertain identification of corresponding networks within and between subjects. MethodsWe adapt a reference-informed network estimation technique to capture time-resolved spatial networks and their dynamic spatial integration and segregation. We focus on time-resolved spatial functional network connectivity (spFNC), an estimate of network spatial coupling, to study sex-specific alterations in schizophrenia and their links to multi-factorial genomic data. ResultsOur findings are consistent with the dysconnectivity and neurodevelopment hypotheses and align with the cerebello-thalamo-cortical, triple-network, and frontoparietal dysconnectivity models, helping to unify them. The potential unification offers a new understanding of the underlying mechanisms. Notably, the posterior default mode/salience spFNC exhibits sex-specific schizophrenia alteration during the state with the highest global network integration and correlates with genetic risk for schizophrenia. This dysfunction is also reflected in high-dimensional (voxel-level) space in regions with weak functional connectivity to corresponding networks. ConclusionsOur method can effectively capture spatially dynamic networks, detect nuanced SZ effects, and reveal the intricate relationship of dynamic information to genomic data. The results also underscore the potential of dynamic spatial dependence and weak connectivity in the clinical landscape.

neuroscience↗

bioRxiv2023

CircRNAs and their regulatory networks associated with the antioxidant enzymes and immune responses of Asian honey bee larvae to fungal infection

Non-coding RNA (ncRNA) plays an important role in the regulation of immune responses, growth, and development in plants and animals. Here, the identification, characteristic investigation, and molecular verification of circRNAs in Apis cerana cerana larval guts were conducted, and the expression pattern of larval circRNAs during Ascosphaera apis infection was analyzed. This was followed by exploration of the potential regulatory part of differentially expressed circRNAs (DEcircRNAs) in host immune responses. A total of 3178 circRNAs in the larval guts of A. c. cerana were identified, with a length distribution ranging from 15 nt to 96007 nt. Additionally, 45, 33, and 48 up-regulated circRNAs, as well as 110, 62, and 38 down-regulated circRNAs were identified in the A.-apis-inoculated 4-, 5-, and 6-day-old larval guts in comparison with the corresponding uninoculated larval guts. These DEcircRNAs were predicted to target 29, 25, and 18 parental genes, which were relative to 12, 20, and 17 GO terms as well as 144, 114, and 61 KEGG pathways, including five cellular and four humoral immune pathways containing melanization, phagosomes, lysosomes, endocytosis, apoptosis, MAPK, Ras, and Jak-STAT signaling pathways. Furthermore, complex competing endogenous RNA (ceRNA) regulatory networks were detected as being formed among DEcircRNAs, DEmiRNAs, and DEmRNAs. The target DEmRNAs were engaged in 36, 47, and 47 GO terms as well as 331, 332, and 331 pathways, including six cellular and six humoral immune-related pathways. In total, nineteen DEcircRNAs, five DEmiRNAs, and three mRNAs were included in the sub-networks relative to three antioxidant enzymes, including superoxide dismutase (SOD), catalase (CAT), and glutathione S-transferase (GST). Finally, back-splicing sites within 15 circRNAs and the difference in the 15 DEcircRNAs expression between uninoculated and A.-apis-inoculated larval guts were confirmed utilizing molecular methods. These findings not only enrich our understanding of bee host-fungal pathogen interactions, but also lay a foundation for illuminating the mechanism underlying the DEcircRNA-mediated immune defense of A. c. cerana larvae against A. apis invasion.

molecular biology↗

bioRxiv2023

A Multimodal Vision Transformer for InterpretableFusion of Functional and Structural NeuroimagingData

Deep learning models, despite their potential for increasing our understanding of intricate neuroimaging data, can be hampered by challenges related to interpretability. Multimodal neuroimaging appears to be a promising approach that allows us to extract supplementary information from various imaging modalities. Its noteworthy that functional brain changes are often more pronounced in schizophrenia, albeit potentially less reproducible, while structural MRI effects are more replicable but usually manifest smaller effects. Instead of conducting isolated analyses for each modality, the joint analysis of these data can bolster the effects and further refine our neurobiological understanding of schizophrenia. This paper introduces a novel deep learning model, the multimodal vision transformer (MultiViT), specifically engineered to enhance the accuracy of classifying schizophrenia by using structural MRI (sMRI) and functional MRI (fMRI) data independently and simultaneously leveraging the combined information from both modalities. This study uses functional network connectivity data derived from a fully automated independent component analysis method as the fMRI features and segmented gray matter volume (GMV) as the sMRI features. These offer sensitive, high-dimensional features for learning from structural and functional MRI data. The resulting MultiViT model is lightweight and robust, outperforming unimodal analyses. Our approach has been applied to data collected from control subjects and patients with schizophrenia, with the MultiViT model achieving an AUC of 0.833, which is significantly higher than the average 0.766 AUC for unimodal baselines and 0.78 AUC for multimodal baselines. Advanced algorithmic approaches for predicting and characterizing these disorders have consistently evolved, though subject and diagnostic heterogeneity pose significant challenges. Given that each modality provides only a partial representation of the brain, we can gather more comprehensive information by harnessing both modalities than by relying on either one independently. Furthermore, we conducted a saliency analysis to gain insights into the co-alterations in structural gray matter and functional network connectivity disrupted in schizophrenia. While its clear that the MultiViT model demonstrates differences compared to previous multimodal methods, the specifics of how it compares to methods such as MCCA and JICA are still under investigation, and more research is needed in this area. The findings underscore the potential of interpretable multimodal data fusion models like the MultiViT, highlighting their robustness and potential in the classification and understanding of schizophrenia.

neuroscience↗

bioRxiv2023

Cryo-EM studies of the four E. coli paralogs establish ABCF proteins as master plumbers of the peptidyl-transferase center of the ribosome

The genomes of most mesophilic organisms encode multiple ATP-Binding Cassette F (ABCF) proteins. EttA, one of four E. coli paralogs, regulates synthesis of the first peptide bond on the ribosome dependent on ATP/ADP ratio, while Antibiotic Resistance factors (AREs), paralogs in other organisms, both regulate and directly mediate resistance to ribosome-targeted antibiotics. However, the physiological functions remain unclear for most paralogs, and the mechanism-of-action has yet to be rigorously established for any paralog. We herein present single particle cryogenic electron microscopy structures of ribosome complexes of all four E. coli ABCF paralogs (EttA, Uup, YbiT, and YheS), which, together with previously determined ARE structures, show that ABCFs control the binding geometry of the tRNA in the peptidyl-tRNA-binding (P) site on the ribosome. They modulate the position of its acceptor stem relative to the peptidyl transferase center (PTC) in a manner that can either promote (EttA and Uup) or disrupt (YbiT, YheS, and the AREs) proper catalytic geometry. The YbiT/70S reconstructions include a conformation with no density for ribosomal protein bL33, and structural analyses support the exchange of this sub-stoichiometric ribosomal protein being functionally related to conformational changes in YbiT controlled by sequence variations in the strongly non-canonical Signature Sequence in its first ABC domain. Our studies establish general structural/enzymological principles by which the ATPase activity of ABCF proteins controls translation elongation coupled to modulation of conformation and stereochemistry in the catalytic core of the ribosome.

biophysics↗

bioRxiv2023

Transfer Learning Enhanced Graph Neural Network for Aldehyde Oxidase Metabolism Prediction and Its Experimental Application

Aldehyde oxidase (AOX) is a molybdoenzyme that is primarily expressed in the liver and is involved in the metabolism of drugs and other xenobiotics. AOX-mediated metabolism can result in unexpected outcomes, such as the production of toxic metabolites and high metabolic clearance, which can lead to the clinical failure of novel therapeutic agents. Computational models can assist medicinal chemists in rapidly evaluating the AOX metabolic risk of compounds during the early phases of drug discovery and provide valuable clues for manipulating AOX-mediated metabolism liability. In this study, we developed a novel graph neural network called AOMP for predicting AOX-mediated metabolism. AOMP integrated the tasks of metabolic substrate/non-substrate classification and metabolic site prediction, while utilizing transfer learning from 13C nuclear magnetic resonance data to enhance its performance on both tasks. AOMP significantly outperformed the benchmark methods in both cross-validation and external testing. Using AOMP, we systematically assessed the AOX-mediated metabolism of common fragments in kinase inhibitors and successfully identified four new scaffolds with AOX metabolism liability, which were validated through in vitro experiments. Furthermore, for the convenience of the community, we established the first online service for AOX metabolism prediction based on AOMP, which is freely available at https://aomp.alphama.com.cn.

bioinformatics↗

bioRxiv2023

BiATNovo: A Self-Attention based Bidirectional Peptide Sequencing Method

De novo sequencing from tandem mass spectra (MS/MS) data is a key technique for identifying novel peptides. In theory, the Data-Independent Acquisition (DIA) method can fragment all precursor ions in an unbiased and non-targeted fashion. However, each spectrum contains fragments from multiple precursor ions, and the unclear relationship between these ions and their fragments poses a significant challenge to the accuracy of de novo sequencing algorithms. Here we present BiATNovo, an attention-based bidirectional de novo peptide sequencing framework. BiATNovo comprises a bidirectional attention-based model and a bidirectional fusion-reranking post-processing module, which enables efficient capture of relationships between tandem mass spectra, fragment ions, and peptide patterns, while also expanding the candidate set to select the optimal sequence. This framework improves peptide prediction accuracy, particularly for long peptide sequences, and mitigates the imbalance where the initial amino acids are predicted more accurately than the last ones. Evaluation results demonstrate that BiATNovo outperforms existing algorithms, including DeepNovo-DIA and PepNet, in both peptid-level and amino acid-level. Furthermore, when extended to DDA datasets, BiATNovo achieves comparable performance to state-of-the-art models.

bioinformatics↗

bioRxiv2023

Cryo-EM of α-tubulin isotype containing microtubules revealed a contracted structure of α4A/β2A microtubules

Microtubules are hollow /{beta}-tubulin heterodimeric polymers playing critical roles in cells. In vertebrates, both - and {beta}-tubulin have multiple isotypes encoded by different genes, which are intrinsic factors in regulating microtubule functions. However, structures of microtubules composed of different tubulin isotypes, especially -tubulin isotypes, remain largely unknown. Here, we purified recombinant tubulin heterodimers composed of different mouse -tubulin isotypes, including 1A, 1C and 4A, with {beta}-tubulin isotype {beta}2A. We further assembled and determined the cryo-electron microscopy (cryo-EM) structures of 1A/{beta}2A, 1C/{beta}2A, and 4A/{beta}2A microtubules. Our structural analysis demonstrated that 4A/{beta}2A microtubules exhibit a longitudinal contraction between tubulin interdimers compared with 1A/{beta}2A and 1C/{beta}2A microtubules. Collectively, our findings reveal that -tubulin isotype composition could tune microtubule structures, and also provide evidence for the "tubulin code" hypothesis.

biophysics↗

bioRxiv2023

Conflicts are represented in a cognitive space to reconcile domain-general and domain-specific cognitive control

Cognitive control resolves conflicts between task-relevant and -irrelevant information to enable goal-directed behavior. As conflicts can arise from different sources (e.g., sensory input, internal representations), how a limited set of cognitive control processes can effectively address diverse conflicts remains a major challenge. Based on the cognitive space theory, different conflicts can be parameterized and represented as distinct points in a (low-dimensional) cognitive space, which can then be resolved by a limited set of cognitive control processes working along the dimensions. It leads to a hypothesis that conflicts similar in their sources are also represented similarly in the cognitive space. We designed a task with five types of conflicts that could be conceptually parameterized. Both human performance and fMRI activity patterns in the right dorsolateral prefrontal (dlPFC) support that different types of conflicts are organized based on their similarity, thus suggesting cognitive space as a principle for representing conflicts.

neuroscience↗

bioRxiv2023

Astrocytic L-lactate signaling in the anterior cingulate cortex is essential for schema memory and neuronal mitochondrial biogenesis

Astrocyte-derived L-lactate was shown to confer beneficial effects on synaptic plasticity and cognitive functions. However, how astrocytic Gi signaling in the anterior cingulate cortex (ACC) modulates L-lactate levels and schema memory is not clear. Here, using chemogenetic approach and well-established behavioral paradigm, we demonstrate that astrocytic Gi pathway activation in ACC causes significant impairment in flavor-place paired associates (PA) learning, schema formation, and PA memory retrieval in rats. It also impairs new PA learning even if a prior associative schema exists. These impairments were mediated by decreased L-lactate in ACC due to astrocytic Gi activation. Concurrent exogenous L-lactate administration bilaterally into the ACC rescues these impairments. Furthermore, we show that the impaired schema memory formation was associated with a decreased neuronal mitochondrial biogenesis caused by decreased L-lactate level in ACC upon Gi activation. Our study also reveals that L-lactate mediated mitochondrial biogenesis is dependent on monocarboxylate transporter 2 and NMDA receptor activity - discovering a previously unrecognized signaling role of L-lactate. These findings expand our understanding of the role of astrocytes and L-lactate in brain functions.

neuroscience↗

bioRxiv2022

Differential Roles of positive and negative supercoiling in compacting the E. coli genome

This study aims to explore whether and how positive and negative supercoiling contribute to the three-dimensional (3D) organization of the bacterial genome. We used recently published Escherichia coli GapR ChIP-seq and TopoI ChIP-seq (also called EcTopoI-seq) data, which marks positive and negative supercoiling sites, respectively, to study how positive and negative supercoiling correlates with the corresponding contact frequencies obtained from chromosome conformation capture sequencing (Hi-C and 5C). We found that supercoiled chromosomal loci have overall higher Hi-C contact frequencies than sites that are not supercoiled, with positive supercoiling surprisingly corresponding to higher spatial contacts than negative supercoiling. Additionally, Hi-C contact frequencies alone could identify positive, but not negative, supercoiling with high accuracy. The majority of positive and negative supercoils coincide with highly active transcription units, with a minor group likely associated with replication and other genomic processes. Our results suggest that both positive and negative supercoiling enhance chromosome interactions, but positive supercoils contribute more than negative supercoils to bring distant chromosomal loci closer in space. Based on these results, we propose new physical models of how the E. coli chromosome is organized differentially by positive and negative supercoils.

bioinformatics↗

bioRxiv2022

Differential expression patterns and potential regulatory function of piRNAs in western honey bee larval response to Ascosphaera apis infestation

piRNAs are a class of small non-coding RNAs that play an essential part in genomic defense, as well as in modulation of gene expression and diverse biological processes such as host-pathogen interaction. However, little is known about the expression pattern of the regulatory function of piRNAs in interactions between honey bees and pathogens. In this work, on the basis of our previously obtained high-quality small RNA-seq datasets from western honey bee (Apis mellifera) larval guts, the differential expression profile of piRNAs in A. mellifera larval guts after Ascosphaera apis infestation was analyzed, followed by structural characterization, target prediction, and regulatory network investigation. The potential roles of differentially expressed piRNAs (DEpiRNAs) in regulating host response, especially immune response, were further discussed. In this study, 504, 657, and 587 piRNAs were respectively identified in the 4-, 5-, and 6-day-old larval guts infected by A. apis, with 411 (53.24%) piRNAs shared. The length distribution of these piRNAs ranged from 24 nt to 33 nt and their first base had a "C" bias, similar to piRNAs discovered in other mammals and insects. Additionally, 96, 103, and 143 DEpiRNAs were detected in the 4-, 5-, and 6-day-old comparison groups; among these, piR-ame-149736, piR-ame-1066173, and piR-ame-1125190 were the most up-regulated, while piR-ame-1202932 was the most down-regulated. There were 68 DEpiRNAs shared between these three comparison groups. The targets of the DEpiRNAs in the three comparison groups were engaged in a suite of crucial functions associated with biological processes, molecular function, and cellular components, including molecular transducer activity, biological regulation, and membrane part. These targets were also relevant to diverse vital pathways such as the phosphatidylinositol signaling system, inositol phosphate metabolism, and Wnt signaling pathway. Further investigation demonstrated that targets of DEpiRNAs were involved in three energy metabolism-related pathways, seven development-associated signaling pathways, and seven immune-relevant pathways, including lysosome and endocytosis, as well as the MAPK and Jak-STAT signaling pathways. The expression trends of five randomly selected DEpiRNAs were verified using a combination of RT-PCR and RT-qPCR. Moreover, the expression levels of six genes targeted by piR-ame-945760 were detected by RT-qPCR, with the results showing that their expression trends were the same as the expression trend of piR-ame-945760, indicative of the positive correlation between piR-ame-945760 and these targets. These results suggest that A.apis infestation increased the overall expression level of piRNAs and altered the expression pattern of piRNAs in A. mellifera larval guts. DEpiRNAs potentially participate in the A. apis response of the host by modulating the expression of target genes associated with energy metabolism and development, as well as cellular and humoral immune response. Our findings not only offer novel insights into A. mellifera larva-A. apis interaction, but also lay the groundwork for clarifying the mechanisms underlying DEpiRNA-regulated larval response.

molecular biology↗

bioRxiv2022

Effect of Ascosphaera apis infestation on the activities of four antioxidant enzymes in Asian honey bee larval guts

Ascosphaera apis exclusively infects bee larvae and causes chalkbrood, a lethal fungal disease that results in the sharp reduction in adult bees and colony productivity. However, little is known about the effect of A. apis infestation on the activities of antioxidant enzymes in bee larvae. Here, A. apis spores were purified and used to inoculate Asian honey bee (Apis cerana) larvae, followed by detection of the host survival rate and evaluation of the activities of four major antioxidant enzymes. At 6 days post inoculation (dpi) with A. apis spores, white mycelia penetrated the posterior end of the larva, extended to the anterior end, and eventually covered the entire larval body surface, presenting an obvious symptom of chalkbrood disease similar to that occurs in Apis mellifera larvae. Additionally, PCR identification showed that the expected fragment was amplified from the A. apis-inoculated larval guts and the A. apis spores, verifying the A. apis infection of A. cerana larvae. The survival rate of larvae inoculated with A. apis was high at 1-2 dpi, sharply decreased to 4.16% at 4 dpi, and reached 0% at 5 dpi; whereas that of un-inoculated larvae was always high at 1~8 dpi, with an average survival rate of 95.37%, indicating the negative impact of A. apis infection on larval survival. Furthermore, in comparison with those in the corresponding un-inoculated groups, the superoxide dismutase (SOD) and catalase (CAT) activities in the 4-day-old larval gut in the A. apis-inoculated groups were reduced (p > 0.05), while those in the 5- and 6-day-old larval guts were significantly decreased (p < 0.05); the glutathione S-transferase (GST) activity in the 4- and 5-day-old larval guts was significantly increased (p < 0.05), while that in the 6-day-old larval gut was reduced (p > 0.05); the polyphenol oxidase (PPO) activity in 4-day-old larval gut was increased (p > 0.05) and that in the 5-day-old larval gut was significantly increased (p < 0.05), whereas that in the 6-day-old larval gut was significantly reduced (p < 0.01). These results together suggested that the activities of SOD and CAT in the larval guts were suppressed during the process of A. apis infestation, while the GST activity was induced to activation, and the PPO activity was first enhanced and then inhibited. Our findings not only unravel the response of A. cerana larvae to A. apis infestation from a biochemical perspective, but also offer a valuable insight into the interaction between Asian honey bee larvae and A. apis.

molecular biology↗

bioRxiv2022

An electrostatic cluster guides Aβ40 fibril formation in cerebral amyloid angiopathy

Cerebral amyloid angiopathy (CAA) is associated with the accumulation of fibrillar A{beta} peptides upon and within the cerebral vasculature, which leads to loss of vascular integrity and contributes to disease progression in Alzheimers disease (AD). We investigate the structure of human-derived A{beta}40 fibrils obtained from patients diagnosed with sporadic or familial Dutch-type (E22Q) CAA. Using cryo-EM, two primary structures are identified containing elements that have not been observed in in vitro A{beta}40 fibril structures. One population has an ordered N-terminal fold comprised of two {beta}-strands stabilized by electrostatic interactions involving D1, E22, D23 and K28. This charged cluster is disrupted in the second population, which exhibits a disordered N-terminus and is favored in fibrils derived from the familial Dutch-type CAA patient. These results illustrate differences between human-derived CAA and AD fibrils, and how familial CAA mutations guide fibril formation.

neuroscience↗

bioRxiv2022

Cis-regulatory evolution of the potassium channel gene kcnj13 during pigment pattern diversification in Danio fish

Teleost fish of the genus Danio are excellent models to study the genetic and cellular bases of pigment pattern variation in vertebrates. The two sister species Danio rerio and Danio aesculapii show divergent patterns of horizontal stripes and vertical bars that are partly caused by the evolution of the potassium channel gene kcnj13. In D. rerio, kcnj13 is required in melanophores for interactions with xanthophores and iridophores, which cause location-specific pigment cell shapes and thereby influence colour pattern and contrast. Here, we show that cis-regulatory rather than protein coding changes underlie kcnj13 evolution between the two species. D. aesculapii express lower kcnj13 levels and exhibit low-contrast patterns similar to D. rerio mutants. Our results suggest that homotypic and heterotypic interactions between the pigment cells and their shapes diverged between species by quantitative changes in kcnj13 expression during pigment pattern diversification.

evolutionary biology↗

bioRxiv2022

Maximum Classifier Discrepancy Generative Adversarial Network for Jointly Harmonizing Scanner Effects and Improving Reproducibility of Downstream Tasks

ObjectiveMulti-site collaboration is essential for overcoming small-sample problems when exploring reproducible biomarkers in MRI studies. However, various scanner-specific factors dramatically reduce the cross-scanner replicability. Moreover, existing harmony methods mostly could not guarantee the improved performance of downstream tasks. Methodswe proposed a new multi-scanner harmony framework, called maximum classifier discrepancy generative adversarial network, or MCD-GAN, for removing scanner effects in the original feature space while preserving substantial biological information for downstream tasks. Specifically, the adversarial generative network was utilized for persisting the structural layout of each sample, and the maximum classifier discrepancy module was introduced for regulating GAN generators by incorporating the downstream tasks. ResultsWe compared the MCD-GAN with other state-of-the-art data harmony approaches (e.g., ComBat, CycleGAN) on simulated data and the Adolescent Brain Cognitive Development (ABCD) dataset. Results demonstrate that MCD-GAN outperformed other approaches in improving cross-scanner classification performance while preserving the anatomical layout of the original images. SignificanceTo the best of our knowledge, the proposed MCD-GAN is the first generative model which incorporates downstream tasks while harmonizing, and is a promising solution for facilitating cross-site reproducibility in various tasks such as classification and regression. The codes of the MCD-GAN are available at https://github.com/trendscenter/MCD-GAN.

bioengineering↗

bioRxiv2022

Functional gradients reveal altered functional segregation in patients with amnestic mild cognitive impairment and Alzheimer' s disease

BackgroundAlzheimers disease (AD) and amnestic mild cognitive impairment (aMCI) are associated with disrupted functional organization in brain networks, yet the finer changes in the topological organization in aMCI and AD remain to be investigated. Connectome gradients are a new tool representing brain functional topological organization in a low-dimensional space to smoothly capture the human macroscale hierarchy. MethodsHere, we examined altered topological organization in aMCI and AD by connectome gradient mapping. We further quantified functional segregation by gradient dispersion. Then, we systematically compared the alterations observed in aMCI and AD patients with those in normal controls (NCs) in a two-dimensional functional gradient space from both the whole-brain level and module level. ResultsCompared with NCs, the first gradient, which described the neocortical hierarchy from unimodal to transmodal regions, showed a distributed and significant suppression in AD patients, while abnormalities were only limited to local regions in aMCI patients. The second gradient showed a decreased pattern in the somatomotor module in both aMCI and AD patients. Furthermore, gradient dispersion showed significant decreases in AD patients at both the global level and module level, whereas this alteration was limited only to limbic areas in aMCI. Notably, we demonstrated that suppressed gradient dispersion in aMCI and AD patients was associated with cognitive scores. ConclusionsChanges in functional gradients could reflect different degrees of altered brain network segregation in aMCI and AD. These findings provide new evidence for altered brain hierarchy in aMCI and AD, which strengthens our understanding of the progressive mechanism of cognitive decline.

neuroscience↗

bioRxiv2022

Multi-model order spatially constrained ICA reveals highly replicable group differences and consistent predictive results from fMRI data

Brain functional networks identified from resting fMRI data have the potential to reveal biomarkers for brain disorders, but studies of complex mental illnesses such as schizophrenia (SZ) often yield mixed results across replication studies. This is likely due in part to the complexity of the disorder, the short data acquisition time, and the limited ability of the approaches for brain imaging data mining. Therefore, the use of analytic approaches which can both capture individual variability while offering comparability across analyses is highly preferred. Fully blind data-driven approaches such as independent component analysis (ICA) are hard to compare across studies, and approaches that use fixed atlas-based regions can have limited sensitivity to individual sensitivity. By contrast, spatially constrained ICA (scICA) provides a hybrid, fully automated solution that can incorporate spatial network priors while also adapting to new subjects. However, scICA has thus far only been used with a single spatial scale. In this work, we present an approach using scICA to extract subject-specific intrinsic connectivity networks (ICNs) from fMRI data at multiple spatial scales (ICA model orders), which also enables us to study interactions across spatial scales. We evaluate this approach using a large N (N>1,600) study of schizophrenia divided into separate validation and replication sets. A multi-scale ICN template was estimated and labeled, then used as input into spatially constrained ICA which was computed on an individual subject level. We then performed a subsequent analysis of multiscale functional network connectivity (msFNC) to evaluate the patient data, including group differences and classification. Results showed highly consistent group differences in msFNC in regions including cerebellum, thalamus, and motor/auditory networks. Importantly, multiple msFNC pairs linking different spatial scales were implicated. We also used the msFNC features as input to a classification model in cross-validated hold-out data and also in an independent test data. Visualization of predictive features was performed by evaluating their feature weights. Finally, we evaluated the relationship of the identified patterns to positive symptoms and found consistent results across datasets. The results verified the robustness of our framework in evaluating brain functional connectivity of schizophrenia at multiple spatial scales, implicated consistent and replicable brain networks, and highlighted a promising approach for leveraging resting fMRI data for brain biomarker development.

neuroscience↗