Discovering individual fingerprints in resting-state functional connectivity using deep neural networks.

Non-negligible idiosyncrasy due to interindividual differences is an ongoing issue in resting-state functional MRI (rfMRI) analysis. We show that a deep neural network (DNN) can be employed for individual identification by learning important features from the time-varying functional connectivity (FC) of rfMRI in the Human Connectome Project. We employed the trained DNN to identify individuals from an independent dataset acquired at our institution. The results revealed that the DNN could successfully identify 300 individuals with an error rate of 2.9% using 15 s time-window and 870 individuals with an error rate of 6.7%. A trained DNN with nonlinear hidden layers led to the proposal of the "fingerprint of FC" (fpFC) as representative edges of individual FC. The fpFCs for individuals exhibited commonly important and individual-specific edges across time-window lengths (from 5 min to 15 s). Furthermore, the utility of our model for another group of subjects was validated, supporting the feasibility of our technique in the context of transfer learning. In conclusion, our study offers an insight into the discovery of the intrinsic mode of the human brain using whole-brain resting-state FC and DNNs.

Neural representations of the perception of handwritten digits and visual objects from a convolutional neural network compared to humans.

We investigated neural representations for visual perception of 10 handwritten digits and six visual objects from a convolutional neural network (CNN) and humans using functional magnetic resonance imaging (fMRI). Once our CNN model was fine-tuned using a pre-trained VGG16 model to recognize the visual stimuli from the digit and object categories, representational similarity analysis (RSA) was conducted using neural activations from fMRI and feature representations from the CNN model across all 16 classes. The encoded neural representation of the CNN model exhibited the hierarchical topography mapping of the human visual system. The feature representations in the lower convolutional (Conv) layers showed greater similarity with the neural representations in the early visual areas and parietal cortices, including the posterior cingulate cortex. The feature representations in the higher Conv layers were encoded in the higher-order visual areas, including the ventral/medial/dorsal stream and middle temporal complex. The neural representations in the classification layers were observed mainly in the ventral stream visual cortex (including the inferior temporal cortex), superior parietal cortex, and prefrontal cortex. There was a surprising similarity between the neural representations from the CNN model and the neural representations for human visual perception in the context of the perception of digits versus objects, particularly in the primary visual and associated areas. This study also illustrates the uniqueness of human visual perception. Unlike the CNN model, the neural representation of digits and objects for humans is more widely distributed across the whole brain, including the frontal and temporal areas.

Reliability of self-reported dispositional mindfulness scales and their association with working memory performance and functional connectivity.

We systematically investigated the link between trait mindfulness scores and functional connectivity (FC) features or behavioral data, to emphasize the importance of the reliability of self-report mindfulness scores. Sixty healthy young male participants underwent two functional MRI runs with three mindfulness or mind-wandering task blocks with an N-back task (NBT) block. The data from 49 participants (age: 23.3 ± 2.8) for whom two sets of the self-reported Mindfulness Attention Awareness Scale (MAAS) and NBT performance were available were analyzed. We divided participants into two groups based on the consistency level of their MAAS scores (i.e., a "consistent" and an "inconsistent" group). Then, the association between the MAAS scores and FC features or NBT performance was investigated using linear regression analysis with p-value correction and bootstrapping. Meaningful associations (a) between MAAS and NBT accuracy (slope = 0.41, CI = [0.10, 0.73], corrected p < 0.05), (b) between MAAS and the FC edges in the frontoparietal network, and (c) between the FC edges and NBT performance were only observed in the consistent group (n = 26). Our findings demonstrate the importance of appropriate screening mechanisms for self-report-based dispositional mindfulness scores when trait mindfulness scores are combined with neuronal features and behavioral data.

Overcoming the imatinib-resistant BCR-ABL mutants with new ureidobenzothiazole chemotypes endowed with potent and broad-spectrum anticancer activity.

The design of kinase inhibitors targeting the oncogenic kinase BCR-ABL constitutes a promising paradigm for treating chronic myeloid leukaemia (CML). Nevertheless, the efficacy of imatinib, the first FDA-approved targeted therapy for CML, is curbed by the emergence of resistance. Herein, we report the identification of the 2-methoxyphenyl ureidobenzothiazole AK-HW-90 (2b) as a potent pan-BCR-ABL inhibitor against imatinib-resistant mutants, particularly T315I. A concise array of six compounds 2a-f was designed based on our previously reported benzothiazole lead AKE-5l to improve its BCR-ABLT315I inhibitory activity. Replacing the 6-oxypicolinamide moiety of AKE-5l with o-methoxyphenyl and changing the propyl spacer with phenyl afforded 2a and AK-HW-90 (2b) with IC50 values of 2.0 and 0.65 nM against BCR-ABLT315I, respectively. AK-HW-90 showed superior anticancer potency to imatinib against multiple cancer cells (NCI), including leukaemia K-562. The obtained outcomes offer AK-HW-90 as a promising candidate for the treatment of CML and other types of cancer.

Metal-Coordinated Phthalocyanines as Platform Molecules for Understanding Isolated Metal Sites in the Electrochemical Reduction of CO2.

Single-atom catalysts (SACs) of non-precious transition metals (TMs) often show unique electrochemical performance, including the electrochemical carbon dioxide reduction reaction (CO2RR). However, the inhomogeneity in their structures makes it difficult to directly compare SACs of different TM for their CO2RR activity, selectivity, and reaction mechanisms. In this study, the comparison of isolated TMs (Fe, Co, Ni, Cu, and Zn) is systematically investigated using a series of crystalline molecular catalysts, namely TM-coordinated phthalocyanines (TM-Pcs), to directly compare the intrinsic role of the TMs with identical local coordination environments on the CO2RR performance. The combined experimental measurements, in situ characterization, and density functional theory calculations of TM-Pc catalysts reveal a TM-dependent CO2RR activity and selectivity, with the free energy difference of ΔG(*HOCO) - ΔG(*CO) being identified as a descriptor for predicting the CO2RR performance.

Quantitative Study of Enantiomer-Specific State Transfer.

We here report on a quantitative study of enantiomer-specific state transfer, performed in a pulsed, supersonic molecular beam. The chiral molecule 1-indanol is cooled to low rotational temperatures (1-2 K) and a selected rotational level in the electronic and vibrational ground state of the most abundant conformer is depleted via optical pumping on the S_{1}←S_{0} transition. Further downstream, three consecutive microwave pulses with mutually perpendicular polarizations and with a well-defined duration and phase are applied. The population in the originally depleted rotational level is subsequently monitored via laser-induced fluorescence detection. This scheme enables a quantitative comparison of experiment and theory for the transfer efficiency in what is the simplest enantiomer-specific state transfer triangle for any chiral molecule, that is, the one involving the absolute ground state level, |J_{K_{a}K_{c}}⟩=|0_{00}⟩. Moreover, this scheme improves the enantiomer enrichment by over an order of magnitude compared to previous works. Starting with a racemic mixture, a straightforward extension of this scheme allows one to create a molecular beam with an enantiomer-pure rotational level, holding great prospects for future spectroscopic and scattering studies.

Immunological and Clinical Phenotyping in Primary Antibody Deficiencies: a Growing Disease Spectrum.

PURPOSE: Although common variable immunodeficiency (CVID) is considered the most prevalent symptomatic primary antibody deficiency (PAD), there is a population with symptomatic PADs that do not meet criteria for CVID. We analyzed clinical and immunological profiles of patients with different PADs to better understand the differences and similarities between CVID and other PADs. METHODS: We extracted clinical and laboratory data of patients with PADs from electronic medical records. Patients were categorized into CVID, IgG subclass 2 deficiency (IgG2D), IgG deficiency (IgGD), and specific antibody deficiency (sAbD) based on basal immunoglobulin levels and pneumococcal vaccine responses. We compared clinical and immunological characteristics in these groups. RESULTS: All patients, regardless of PAD types, showed similar frequencies of infections, bronchiectasis, and interstitial lung disease (ILD). Hematopoietic malignancies were more frequently found in the CVID than in the IgG2D, IgGD, and sAbD groups, while the latter groups trended towards an increased frequency of connective tissue diseases (CTD). Low counts of natural killer (NK) cells were associated with malignancy, autoimmunity, and ILD in CVID but not in other PAD groups. CONCLUSIONS: Higher frequency of hematopoietic malignancy in CVID than in the other PADs and association of lower NK cell counts with non-infectious complications in CVID suggest a relationship between immune alterations and the development of non-infectious manifestations in PADs.

Mediation analysis of triple networks revealed functional feature of mindfulness from real-time fMRI neurofeedback.

The triple networks, namely the default-mode network (DMN), the central executive network (CEN), and the salience network (SN), play crucial roles in disorders of the brain, as well as in basic neuroscientific processes such as mindfulness. However, currently, there is no consensus on the underlying functional features of the triple networks associated with mindfulness. In this study, we tested the hypothesis that (a) the partial regression coefficient (i.e., slope): from the SN to the DMN, mediated by the CEN, would be one of the potential mindfulness features in the real-time functional magnetic resonance imaging (rtfMRI) neurofeedback (NF) setting, and (b) this slope level may be enhanced by rtfMRI-NF training. Sixty healthy mindfulness-naïve males participated in an MRI session consisting of two non-rtfMRI-runs, followed by two rtfMRI-NF runs and one transfer run. Once the regions-of-interest of each of the triple networks were defined using the non-rtfMRI-runs, the slope level was calculated by mediation analysis and used as neurofeedback information, in the form of a thermometer bar, to assist with participant mindfulness during the rtfMRI-NF runs. The participants were asked to increase the level of the thermometer bar while deploying a mindfulness strategy, which consisted of focusing attention on the physical sensations of breathing. rtfMRI-NF training was conducted as part of a randomized controlled trial design, in which participants were randomly assigned to either an experimental group or a control group. The participants in the experimental group received contingent neurofeedback information, which was obtained from their own brain signals, whereas the participants in the control group received non-contingent neurofeedback information that originated from matched participants in the experimental group. Our results indicated that the slope level from the SN to the DMN, mediated by the CEN, was associated with mindfulness score (rtfMRI-NF runs: r = 0.53, p = 0.007; p-value was corrected from 10,000 random permutations) and with task-performance feedback score (rtfMRI-NF run: r = 0.61, p = 0.001) in the experimental group only. In addition, during the rtfMRI-NF runs the level of the partial regression coefficient feature was substantially increased in the experimental group compared to the control group (p < 0.05 from the paired t-test; the p-value was corrected from 10,000 random permutations). To the best of our knowledge, this is the first study to demonstrate a partial regression coefficient feature of mindfulness in the rtfMRI-NF setting obtained by triple network mediation analysis, as well as the possibility of enhancement of the partial regression coefficient feature by rtfMRI-NF training.

Novel 5,6-disubstituted pyrrolo[2,3-d]pyrimidine derivatives as broad spectrum antiproliferative agents: Synthesis, cell based assays, kinase profile and molecular docking study.

Two new series of 5-subtituted and 5,6-disubstituted pyrrolo[2,3-d]pyrimidine octamides (4a-o and 6a-g) and their corresponding free amines 5a-m and 7a-g have been synthesized and biologically evaluated for their antiproliferative activity against three human cancer cell lines. The 5,6-disubstituted octamides 6d-g as well as the amine derivative 7b have shown the best anticancer activity with single digit micromolar GI50 values over the tested cancer cells, and low cytotoxic effects (GI50 > 10.0 µM) against HFF-1 normal cell. A structure activity relationship (SAR) study has been established and disclosed that terminal octamide moiety at C2 as well as disubstitution with fluorobenzyl piperazines at C5 and C6 of pyrrolo[2,3-d]pyrimidine are the key structural features prerequisite for best antiproliferative activity. Moreover, the most active member 6f was tested for its antiproliferative activity over a panel of 60 cancer cell lines at NCI, and exhibited distinct broad spectrum anticancer activity with submicromolar GI50 and TGI values over multiple cancer cells. Kinase profile of compound 6f over 53 oncogenic kinases at 10 µM concentration showed its highly selective inhibitory activity towards FGFR4, Tie2 and TrkA kinases. The observed activity of 6f against TrkA (IC50 = 2.25 µM), FGFR4 (IC50 = 6.71 µM) and Tie2 (IC50 = 6.84 µM) was explained by molecular docking study, which also proposed that 6f may be a type III kinase inhibitor, binding to an allosteric site rather than kinase hinge region. Overall, compound 6f may serve as a promising anticancer lead compound that could be further optimized for development of potent anticancer agents.

Synthesis and in vitro antiproliferative activity of C5-benzyl substituted 2-amino-pyrrolo[2,3-d]pyrimidines as potent Hsp90 inhibitors.

A novel series of heat shock protein 90 (Hsp90) inhibitors was identified by X-ray crystal analysis of complex structures at solvent-exposed exit pocket C. The 2-amino-pyrrolo[2,3-d]pyrimidine derivatives, 7-deazapurines substituted with a benzyl moiety at C5, showed potent Hsp90 inhibition and broad-spectrum antiproliferative activity against NCI-60 cancer cell lines. The most potent compound, 6a, inhibited Hsp90 with an IC50 of 36nM and showed a submicromolar mean GI50 value against NCI-60 cell lines. The interaction of 6a at the ATP-binding pocket of Hsp90 was confirmed by X-ray crystallography and Western blot analysis.