Previous evidence points to the benefit of early literacy intervention and support for bilingual children. Therefore, the need for culturally responsive practices in the field of speech-language pathology is essential for the growing bilingual population across all settings. Although there is substantial literature on the home literacy environment (i.e., resources and practices that families use during book reading at home), most of the existing evidence comes from monolingual children. The purpose of our research is to analyze the important relationship between child and caregiver during shared book reading to understand parents' language use and its impact on child vocabulary and grammar. Our research questions are: 1) what is the amount of book reading reported by families; 2) how much bilingual input does the parent produce; and 3) how is this related to children's vocabulary and grammatical development? We collected videos of parent-child book reading in the home and we measured vocabulary development with a parent-report vocabulary checklist in English and Spanish. We also collected environmental surveys containing questions regarding the literacy environment (e.g., duration of book reading; how often they read; and what language they read in). We hypothesize that children who receive more bilingual input during book reading will demonstrate larger vocabularies and more complex syntax. As research assistants in the Child Language Lab, we score standardized language assessments and transcribe the book reading interactions. We have completed scoring and are in the process of analyzing the data from parent report instruments and the transcribed videos. This research may identify key factors in creating an enriching, supportive literacy and language environment for developing multilingual speakers. We can use the findings of this study in the field of Communication Sciences and Disorders and may improve interventions for bilingual children, especially for daycares and preschools. 

Soluble peptides from the HIV-1 (Human Immunodeficiency Virus) envelope heptad repeat-2 domain, known as HIV fusion inhibitors, can inhibit viral entry by blocking formation of the gp41 6-helix bundle required for membrane fusion and infection. However, this treatment is unfeasible because it requires twice-daily subcutaneous injections with high risk and cost. The Lieber Lab is working to engineer hematopoietic stem and progenitor cells (HSPCs) to express HIV fusion inhibitors in vivo, potentially offering sustained protection against HIV. In my work I used SIVmac239 (Simian Immunodeficiency Virus) challenged Rhesus Macaques sera and developed viremia (from another study by Lieber lab). My goal was to test whether anti-gp41 antibodies from these animals cross-reacted with synthetic gp41-derived fusion inhibitor peptides, specifically C46-v2o, C34-SFT, and Enfuvirtide(T20). If antibodies interfered with fusion inhibitors, their therapeutic effect would be severely compromised. In my project, I developed an Enzyme-Linked Immunosorbent Assay (ELISA) to measure antibody titers. These peptides were coated, then blocked with 3% bovine serum albumin, and incubated with diluted Macaque serum to allow antibody binding. I used anti-monkey immunoglobulin-G conjugated with Horseradish Peroxidase for detection of antibody binding. I optimized the serum dilution to 1:200 to reduce background signal and concluded SIV-challenged Macaques had detectable antibody levels against C46-v2o and C34-SFT, but not T20. Ongoing work will determine more detailed IC50 antibody titers in serum samples. Notably, animals with high viral loads exhibited higher levels of antibodies against HIV fusion inhibitors. T20 is a promising candidate for sustained HIV inhibition, as no detectable antibodies means it’s less susceptible to pre-existing immune responses. These findings provide valuable insights into how fusion inhibitors interact with the immune system and help refine strategies for HSPC-based HIV therapies, bringing us closer to a long-term, self-sustaining approach for HIV prevention. 
 

Global interest in and demand for macroalgae farms has increased due to growing interest in seaweeds as food, biofuel, a possible source for carbon sequestration, and an economic asset for small island communities. However, the presence of farms can negatively impact marine mammals, with risks including entanglement, habitat exclusion, and behavioral changes. In this study, we conducted passive acoustic monitoring over three years (2021-2024) in developing macroalgae 5-line and catenary farms at Romero and Media Luna reefs off the southwest coast near La Parguera, Puerto Rico, to assess the long-term effect on bottlenose dolphin (Tursiops truncatus) movement and behavior in the region. Detections were automated using PAMguard software and manually confirmed with visual validation by trained acousticians. We observe higher interaction with farms at dawn and dusk, which corresponds with the crepuscular nature of T. truncatus. To date, we observe no entanglements, and our models indicate no significant long-term effect on the local bottlenose dolphins at low macroalgae biomass. As biomass increases, this may change. This research highlights the important trade-offs between economic development and conservation and the promise of passive acoustic monitoring as an effective, data-rich tool for managers. 

Dense-core vesicles are membrane-bound structures that carry neuromodulators such as insulin, dopamine, and serotonin. The peptides within dense-core vesicles are initially larger precursor proteins that undergo enzymatic processing to achieve their functional forms. During the defecation motor program in Caenorhabditis elegans, dense-core vesicles released from the intestine harbor neuropeptides that trigger neurons which activate enteric muscles, promoting the act of defecation. Failure of certain proneuropeptides to mature into neuropeptides results in decreased frequency of defecations. CPD-1, a conserved transmembrane carboxypeptidase, is a poorly understood processing enzyme that affects the defecation motor program. I built on our knowledge of EGL-21, another carboxypeptidase known to process neuropeptides and peptide hormones, to better understand CPD-1’s function. I show here that these two carboxypeptidases, EGL-21 and CPD-1, process neuropeptides necessary for successful defecation patterns. Mutants lacking egl-21 had decreased defecation frequency while worms lacking both egl-21 and cpd-1 had an even lower defecation frequency. Additionally, my results show that CPD-1 is expressed in intestinal cells and can compensate for EGL-21’s function. Finally, I am conducting experiments to determine whether one of CPD-1’s targets is NLP-40, an important neuro-like peptide released from the intestine that regulates defecation. These results contribute to our broader knowledge of peptide processing in dense-core vesicles.

Adaptive algorithms used in brain-computer interfaces (BCIs) adjust to user strategies by dynamically adjusting how BCIs decode neural data throughout an experiment. Current adaptive algorithms continuously update the decoder using all available data during training. However, if users are unfocused or inattentive, it is likely that some of the training data is unhelpful towards decoder training and could lead to poor decoder performance. Unfortunately, determining attentiveness in a subject is difficult. Non-human animals cannot self-report attention levels, and even in human trials, self-reporting often leads to subjective data that varies between subjects. A non-invasive estimate of subject attentiveness could improve data selection for decoder training. Pupil size is correlated with a participant’s perception of task difficulty, and participants involved in attention-grabbing tasks display blinking rate-inhibition (Kucewicz et al., 2018; Maffei et al., 2019). I hypothesize that these eye data could be used to estimate a subject’s task engagement. I explored data from a novel task where both human and non-human primate subjects controlled a cursor on a 2D screen with 3D hand motions through some unknown mapping. Due to the unknown 3D-to-2D mapping, this task required constant cognitive attention in order for subjects to succeed. I looked at the subjects’ data to identify trends in pupil size and blink frequency across multiple days of task performance. In the future, I aim to build engagement classification models to better select training data for adaptive algorithms and apply these algorithms to realtime BCI experiments.

Brain-computer interfaces (BCIs) decode neural signals from the motor cortex to enable direct control of external devices. While existing BCI designs often combine signals from the premotor (PMd) and primary motor (M1) cortices, these regions have distinct functional roles and anatomical organizations. Prior research demonstrates that PMd and M1 play distinct roles in movement preparation and execution, with information generally flowing from PMd to M1 (Cisek & Kalaska, 2005). Additionally, cortical processing is known to occur in a layer-dependent manner (Bastos et al., 2012), suggesting that different depths within these motor areas may encode distinct aspects of task-related information, highlighting the need for depth-specific analyses. My hypothesis is that task-related information flows directionally from deeper layers of PMd to superficial layers of M1 as behavior transitions from movement preparation to execution. To investigate this, I used Neuropixel probes, which provide high-resolution sampling of neural activity across cortical depths, and performed simultaneous PMd and M1 recordings in two male rhesus macaques as they performed an arm reaching (center-out) task. Preliminary analyses provide evidence that (1) different cortical depths in PMd and M1 encode distinct movement-related and planning information, (2) neural activity in deep PMd exhibits stronger coherence with superficial M1 compared to other depth pairings within and across regions, particularly during movement-related periods, and (3) information flow between PMd and M1 is depth and directionally organized, with information flowing from deep layers of PMd to superficial layers of M1. These findings suggest that the spatial and temporal dynamics of task-related information across cortical depths are important for motor control. Revealing how task-related signals are organized and transmitted across motor cortical layers can inform the development of BCIs that target recordings to leverage these functional dynamics.

Most real-world motor tasks involve a many-to-few input-output relationship, such as many neurons firing or muscles contracting to control a few degrees of freedom of the arm. The brain must form an internal model of outputs to inputs when there are fewer dimensions of feedback than dimensions of inputs to control ("redundancy"). However, motor learning is typically studied in laboratory contexts with one-to-one input-output tasks (Krakauer et al., 2019). To investigate how redundancy influences motor learning, I developed a novel virtual reality (VR) visuomotor perturbation task that can either be fully dimensioned or redundant. Participants are trained on a point-to-point reaching task controlled by hand movements in 3D space. In the 3D version of the task, the 3D cursor motion results from the 3D movements of the hand according to some unknown spatial rotation that the participant will have to learn in order to get to targets. In the redundant (2D) version, 3D hand motions are projected onto a 2D plane oriented somewhere in space that the participant has to learn. In both cases, targets are represented as infinite cylinders such that there is a task-irrelevant dimension, but in the redundant version of the task, the participant receives no visual feedback from this dimension. I hypothesize that providing 3D cursor feedback will enhance the learning of the task-relevant 2D plane by allowing participants to better map redundant hand movements in 3D space onto the constrained 2D plane. In contrast, restricting feedback to only the 2D plane will limit available sensory information, making it more difficult to learn the correct movement strategy. By comparing performance across these two tasks, I aim to clarify how task redundancy influences internal model formation and adaptation, with implications for designing more effective motor rehabilitation and VR-based training protocols.