The Application of Knowledge Engineering via the Use of a Biomimetic Digital Twin Ecosystem, Phenotype-Driven Variant Analysis, and Exome Sequencing to Understand the Molecular Mechanisms of Disease.

Applied artificial intelligence, particularly large language models, in biomedical research is accelerating, but effective discovery and validation requires a toolset without limitations or bias. On January 30, 2023, the National Academies of Sciences, Engineering, and Medicine (NAS) appointed an ad hoc committee to identify the needs and opportunities to advance the mathematical, statistical, and computational foundations of digital twins in applications across science, medicine, engineering, and society. On December 15, 2023, the NAS released a 164-page report, "Foundational Research Gaps and Future Directions for Digital Twins." This report described the importance of using digital twins in biomedical research. The current study was designed to develop an innovative method that incorporated phenotype-ranking algorithms with knowledge engineering via a biomimetic digital twin ecosystem. This ecosystem applied real-world reasoning principles to nonnormalized, raw data to identify hidden or "dark" data. Clinical exome sequencing study on patients with endometriosis indicated four variants of unknown clinical significance potentially associated with endometriosis-related disorders in nearly all patients analyzed. One variant of unknown clinical significance was identified in all patient samples and could be a biomarker for diagnostics. To the best of our knowledge, this is the first study to incorporate the recommendations of the NAS to biomedical research. This method can be used to understand the mechanisms of any disease, for virtual clinical trials, and to identify effective new therapies.

ERPs in Children and Adolescents with Generalized Anxiety Disorder: Before and after an Intervention Program.

According to DSM 5, generalized anxiety disorder (GAD) is characterized by excessive, uncontrollable worry about various topics that occupies the majority of the subject's time for a period of at least six months. The aforementioned state causes distress and/or functional impairments. This paper presents the outcomes of a pilot study that evaluated the implementation of cognitive behavioral therapy (CBT) and CBT with an SSRIs intervention program. The participants comprised 16 children and adolescents with GAD (8 males and 8 females) matched with 16 typically developing peers (8 males and 8 females) aged from 10 to 16 years old (M = 12.56 SD = 2.18). Baseline assessment consisted of event related potentials (ERPs), which indicated that participants with GAD presented cognitive deficits in attention and memory, as they exhibited longer P300 latencies. Following treatment with the CBT program and/or medication, children and adolescents with GAD did not present statistically significantly longer P300 latencies and reaction times in comparison to the control group. Lastly, children and adolescents who followed the CBT program or the CBT program with psychopharmacological assistance did not reveal statistically significant differences in 13 out of 15 topographic brain areas and in reaction time.

Design Space Exploration of a Sparse MobileNetV2 Using High-Level Synthesis and Sparse Matrix Techniques on FPGAs.

Convolution Neural Networks (CNNs) are gaining ground in deep learning and Artificial Intelligence (AI) domains, and they can benefit from rapid prototyping in order to produce efficient and low-power hardware designs. The inference process of a Deep Neural Network (DNN) is considered a computationally intensive process that requires hardware accelerators to operate in real-world scenarios due to the low latency requirements of real-time applications. As a result, High-Level Synthesis (HLS) tools are gaining popularity since they provide attractive ways to reduce design time complexity directly in register transfer level (RTL). In this paper, we implement a MobileNetV2 model using a state-of-the-art HLS tool in order to conduct a design space exploration and to provide insights on complex hardware designs which are tailored for DNN inference. Our goal is to combine design methodologies with sparsification techniques to produce hardware accelerators that achieve comparable error metrics within the same order of magnitude with the corresponding state-of-the-art systems while also significantly reducing the inference latency and resource utilization. Toward this end, we apply sparse matrix techniques on a MobileNetV2 model for efficient data representation, and we evaluate our designs in two different weight pruning approaches. Experimental results are evaluated with respect to the CIFAR-10 data set using several different design methodologies in order to fully explore their effects on the performance of the model under examination.

Single cell transcriptome in aneuploidies reveals mechanisms of gene dosage imbalance.

Aneuploidy is a major source of gene dosage imbalance due to copy number alterations (CNA), and viable human trisomies are model disorders of altered gene expression. We study gene and allele-specific expression (ASE) of 9668 single-cell fibroblasts from trisomy 21 (T21) discordant twins and from mosaic T21, T18, T13 and T8. We examine 928 single cells with deep scRNAseq. Expected and observed overexpression of trisomic genes in trisomic vs. diploid bulk RNAseq is not detectable in trisomic vs. diploid single cells. Instead, for trisomic genes with low-to-average expression, their altered gene dosage is mainly due to the higher fraction of trisomic cells simultaneously expressing these genes, in agreement with a stochastic 2-state burst-like model of transcription. These results, confirmed in a further analysis of 8740 single fibroblasts with shallow scRNAseq, suggest that the specific transcriptional profile of each gene contributes to the phenotypic variability of trisomies. We propose an improved model to understand the effects of CNA and, generally, of gene regulation on gene dosage imbalance.

Extensive cellular heterogeneity of X inactivation revealed by single-cell allele-specific expression in human fibroblasts.

X-chromosome inactivation (XCI) provides a dosage compensation mechanism where, in each female cell, one of the two X chromosomes is randomly silenced. However, some genes on the inactive X chromosome and outside the pseudoautosomal regions escape from XCI and are expressed from both alleles (escapees). We investigated XCI at single-cell resolution combining deep single-cell RNA sequencing with whole-genome sequencing to examine allelic-specific expression in 935 primary fibroblast and 48 lymphoblastoid single cells from five female individuals. In this framework we integrated an original method to identify and exclude doublets of cells. In fibroblast cells, we have identified 55 genes as escapees including five undescribed escapee genes. Moreover, we observed that all genes exhibit a variable propensity to escape XCI in each cell and cell type and that each cell displays a distinct expression profile of the escapee genes. A metric, the Inactivation Score-defined as the mean of the allelic expression profiles of the escapees per cell-enables us to discover a heterogeneous and continuous degree of cellular XCI with extremes represented by "inactive" cells, i.e., cells exclusively expressing the escaping genes from the active X chromosome and "escaping" cells expressing the escapees from both alleles. We found that this effect is associated with cell-cycle phases and, independently, with the XIST expression level, which is higher in the quiescent phase (G0). Single-cell allele-specific expression is a powerful tool to identify novel escapees in different tissues and provide evidence of an unexpected cellular heterogeneity of XCI.

On Unrooted and Root-Uncertain Variants of Several Well-Known Phylogenetic Network Problems.

The hybridization number problem requires us to embed a set of binary rooted phylogenetic trees into a binary rooted phylogenetic network such that the number of nodes with indegree two is minimized. However, from a biological point of view accurately inferring the root location in a phylogenetic tree is notoriously difficult and poor root placement can artificially inflate the hybridization number. To this end we study a number of relaxed variants of this problem. We start by showing that the fundamental problem of determining whether an unrooted phylogenetic network displays (i.e. embeds) an unrooted phylogenetic tree, is NP-hard. On the positive side we show that this problem is FPT in reticulation number. In the rooted case the corresponding FPT result is trivial, but here we require more subtle argumentation. Next we show that the hybridization number problem for unrooted networks (when given two unrooted trees) is equivalent to the problem of computing the tree bisection and reconnect distance of the two unrooted trees. In the third part of the paper we consider the "root uncertain" variant of hybridization number. Here we are free to choose the root location in each of a set of unrooted input trees such that the hybridization number of the resulting rooted trees is minimized. On the negative side we show that this problem is APX-hard. On the positive side, we show that the problem is FPT in the hybridization number, via kernelization, for any number of input trees.

Detection of Imprinted Genes by Single-Cell Allele-Specific Gene Expression.

Genomic imprinting results in parental-specific gene expression. Imprinted genes are involved in the etiology of rare syndromes and have been associated with common diseases such as diabetes and cancer. Standard RNA bulk cell sequencing applied to whole-tissue samples has been used to detect imprinted genes in human and mouse models. However, lowly expressed genes cannot be detected by using RNA bulk approaches. Here, we report an original and robust method that combines single-cell RNA-seq and whole-genome sequencing into an optimized statistical framework to analyze genomic imprinting in specific cell types and in different individuals. Using samples from the probands of 2 family trios and 3 unrelated individuals, 1,084 individual primary fibroblasts were RNA sequenced and more than 700,000 informative heterozygous single-nucleotide variations (SNVs) were genotyped. The allele-specific coverage per gene of each SNV in each single cell was used to fit a beta-binomial distribution to model the likelihood of a gene being expressed from one and the same allele. Genes presenting a significant aggregate allelic ratio (between 0.9 and 1) were retained to identify of the allelic parent of origin. Our approach allowed us to validate the imprinting status of all of the known imprinted genes expressed in fibroblasts and the discovery of nine putative imprinted genes, thereby demonstrating the advantages of single-cell over bulk RNA-seq to identify imprinted genes. The proposed single-cell methodology is a powerful tool for establishing a cell type-specific map of genomic imprinting.

A large genomic deletion leads to enhancer adoption by the lamin B1 gene: a second path to autosomal dominant adult-onset demyelinating leukodystrophy (ADLD).

Chromosomal rearrangements with duplication of the lamin B1 (LMNB1) gene underlie autosomal dominant adult-onset demyelinating leukodystrophy (ADLD), a rare neurological disorder in which overexpression of LMNB1 causes progressive central nervous system demyelination. However, we previously reported an ADLD family (ADLD-1-TO) without evidence of duplication or other mutation in LMNB1 despite linkage to the LMNB1 locus and lamin B1 overexpression. By custom array-CGH, we further investigated this family and report here that patients carry a large (∼660 kb) heterozygous deletion that begins 66 kb upstream of the LMNB1 promoter. Lamin B1 overexpression was confirmed in further ADLD-1-TO tissues and in a postmortem brain sample, where lamin B1 was increased in the frontal lobe. Through parallel studies, we investigated both loss of genetic material and chromosomal rearrangement as possible causes of LMNB1 overexpression, and found that ADLD-1-TO plausibly results from an enhancer adoption mechanism. The deletion eliminates a genome topological domain boundary, allowing normally forbidden interactions between at least three forebrain-directed enhancers and the LMNB1 promoter, in line with the observed mainly cerebral localization of lamin B1 overexpression and myelin degeneration. This second route to LMNB1 overexpression and ADLD is a new example of the relevance of regulatory landscape modifications in determining Mendelian phenotypes.

Diagnostic exome sequencing to elucidate the genetic basis of likely recessive disorders in consanguineous families.

Rare, atypical, and undiagnosed autosomal-recessive disorders frequently occur in the offspring of consanguineous couples. Current routine diagnostic genetic tests fail to establish a diagnosis in many cases. We employed exome sequencing to identify the underlying molecular defects in patients with unresolved but putatively autosomal-recessive disorders in consanguineous families and postulated that the pathogenic variants would reside within homozygous regions. Fifty consanguineous families participated in the study, with a wide spectrum of clinical phenotypes suggestive of autosomal-recessive inheritance, but with no definitive molecular diagnosis. DNA samples from the patient(s), unaffected sibling(s), and the parents were genotyped with a 720K SNP array. Exome sequencing and array CGH (comparative genomic hybridization) were then performed on one affected individual per family. High-confidence pathogenic variants were found in homozygosity in known disease-causing genes in 18 families (36%) (one by array CGH and 17 by exome sequencing), accounting for the clinical phenotype in whole or in part. In the remainder of the families, no causative variant in a known pathogenic gene was identified. Our study shows that exome sequencing, in addition to being a powerful diagnostic tool, promises to rapidly expand our knowledge of rare genetic Mendelian disorders and can be used to establish more detailed causative links between mutant genotypes and clinical phenotypes.