Browse Articles

Challenges and realities of early childhood development centers in Malawi: A critical examination

The study on Early Childhood Development (ECD) practices in T/A Zilakoma, Nkhata Bay South Constituency, Malawi, employs the Ecological Systems Theory to explore recruitment, role definitions, and support systems. This theoretical construct enables an intricate examination of interactions within various environmental systems, emphasizing micro, meso, exo, macro, and chrono systems. Specifically, it illuminates the dynamics within immediate settings, interconnections among diverse systems, broader indirect influences, cultural ideologies, societal values, and temporal dimensions, offering a comprehensive lens for understanding educational contexts. The descriptive qualitative research was conducted within a case study framework to explore the practical experiences of stakeholders within ECD using semi-structured interview guides. Ethical standards were upheld, ensuring voluntary participation and confidentiality. Purposive sampling was used to collect data from diverse and knowledgeable participants involved in ECD domains, providing comprehensive insights aligned with the study’s objectives. Thematic analysis and sentiment mining were performed using Atlas 23 software. The results revealed themes such as recruitment practices relying on community-driven approaches, role ambiguity due to undefined responsibilities, informal evaluation processes, inconsistent training opportunities, and a dependency on community and volunteerism. These themes highlight the absence of formal structures and standardized processes in various aspects of ECD programs. Additionally, sentiment analysis illustrated diverse perspectives among stakeholders, reflecting their distinct experiences and challenges within the ECD landscape. The study concludes with policy recommendations aimed at addressing these systemic challenges.

Behind the mask: Random and selective masking in transformer models applied to specialized social science texts

Transformer models such as BERT and RoBERTa are increasingly popular in the social sciences to generate data through supervised text classification. These models can be further trained through Masked Language Modeling (MLM) to increase performance in specialized applications. MLM uses a default masking rate of 15 percent, and few works have investigated how different masking rates may affect performance. Importantly, there are no systematic tests on whether selectively masking certain words improves classifier accuracy. In this article, we further train a set of models to classify fake news around the coronavirus pandemic using 15, 25, 40, 60 and 80 percent random and selective masking. We find that a masking rate of 40 percent, both random and selective, improves within-category performance but has little impact on overall performance. This finding has important implications for scholars looking to build BERT and RoBERTa classifiers, especially those where one specific category is more relevant to their research.

Isolation and characterization of bacteriophage against clinical isolates of AmpC beta lactamase–Producing Klebsiella pneumoniae from hospital wastewater

Background The increasing incidence of AmpC β-lactamase producing by K. pneumoniae has raised global alarm. Consequently, there is a crucial need for effective methods to inactivate pathogenic bacteria and mitigate the associated risks. Bacteriophage therapy has been demonstrated to be an effective and alternative approach for targeting and inactivating K. pneumoniae that produces AmpC. This study aimed to isolate and characterize the Klebsiella pneumoniae AmpC-specific phages from hospital wastewater. Methods The hospital wastewater samples were collected from the sewage water effluent of a tertiary hospital at Universiti Sains Malaysia, located on the east coast of Malaysia. These samples underwent serial filtration and centrifugation processes for phage recovery. The phage solutions were undergoing a screening test by spot assay using clinical isolates of Klebsiella pneumoniae AmpC strain as amplification hosts. The isolated AmpC-phages were further studied and characterised to determine the phage’s host range, temperature, pH, and chloroform stabilities. High-Resolution Transmission Electron Microscopy (HRTEM) was performed to determine the phage type. Results Thirty HWW samples were analyzed using four K. pneumoniae AmpC strains resulting in a total of 120 screening plates. The AmpC—Klebsiella pneumoniae (AmpC-KP) phages were detected in 31.70% (38/120) of the plates. The AmpC-KP phages had lytic diameters ranging from 1–3 mm, and a phage titer ranged from4×103–3.2×107 PFU/ml. The phages had a narrow–host range stable at a temperature range from -20–50˚C. The phages were also stable at pH ranging from 4 to 9 and at different concentrations of chloroform (5%,10%). Based on HRTEM, Siphoviridea was identified. Conclusions The AmpC-phages were abundant in hospital wastewater, and HWW was a good source for AmpC-KP phages. The isolated AmpC phages had a high effectivity and specificity for AmpC-KP with a narrow host range and could survive under harsh conditions such as (temperature, pH, and chloroform).

Factors associated with phosphate homeostasis in children with beta-thalassemia major: An analytical cross sectional study from Pakistan

Introduction Children with beta-thalassemia major (β-TM) commonly experience metabolic bone diseases. Understanding fibroblast growth factor 23 (FGF-23) levels in these children can shed light on phosphate dysregulation. This study aimed to assess changes in phosphate homeostasis and associated factors, including FGF-23 and explore relationships between iron overload, FGF23 levels, and phosphorus regulation for clinical management of phosphate disorders, in children with β-TM. Methods 143 β-TM patients (57.3% male, median age 12 years) were recruited from Fatimid Foundation Karachi, a blood transfusion facility from January to October 2022. Clinical and biochemical evaluations were conducted at Aga Khan University Hospital, including serum ferritin, calcium (Ca), phosphate (P), vitamin D levels, and FGF-23. Descriptive and inferential statistics including multivariable analysis were applied. Results This study enrolled 143 patients, with 57.3% males. The median age was 12 years, with 53% underweight. Blood transfusion rates varied, with 66.4% receiving 2/month. Bone/joint pain was reported by 76.2%, with 60.8% requiring analgesics. Median serum ferritin was 2768.3 ng/mL. Hypophosphatemia and hyperphosphatemia were observed in 5.6% and 3.5% of participants, respectively. Vitamin D deficiency/insufficiency affected 92.3%. Plasma c-FGF23 was elevated in 60.8%, while i-FGF23 was high in 14%. A low TMP-GFR (glomerular filtration rate) was associated with high c-FGF23 and low i-FGF23. Multivariable regression revealed c-FGF23, TMP:GFR, Corrected Ca, iPTH, and an interaction term between corrected Ca and iPTH as predictors of serum P variability (~75%). Conclusion The study identified contributors to the variations observed in serum P levels in individuals with β-TM and recommends multidisciplinary care and prospective future studies to form targeted interventions for this population.

An efficient Rhizobium rhizogenes-mediated transformation system for Cuscuta campestris

Parasitism has evolved independently in various plant families, with Cuscuta campestris (field dodder) being an economically significant example. Despite advances in genomics and transcriptomics, functional studies in C. campestris are limited by the lack of an efficient genetic transformation system. This study introduces a highly effective Rhizobium rhizogenes-mediated transformation system for C. campestris using a pBIN plasmid harboring a Yellow Fluorescence Protein reporter gene. We optimized transformation and regeneration by assessing explant type, media composition, and plant growth regulators. Notably, host plant contact was essential for transgenic shoot regeneration. Over 70% transformation efficiency was achieved using cuttings co-incubated with modified Murashige and Skoog medium and 5 mg/L Benzylaminopurine, followed by transfer to tomato hosts. Additionally, we developed a complete in-vivo protocol over 30% regeneration efficiency. Transgenic shoots were confirmed for rol gene expression and haustoria formation, advancing functional studies in C. campestris.

Characterization of fillets and skins from two varieties of genetically improved farmed Nile tilapia (Oreochromis niloticus)

Genetically improved farmed tilapia (GIFT) is a strain of Nile tilapia (Oreochromis niloticus) developed for improved production and commercial parameters. Skin color, one of the characteristics distinguishing tilapia varieties, is an important phenotypic trait associated with qualitative and productive performance. This study aimed to assess fillet and skin characteristics in GIFT Nile tilapia with black and red skin phenotypes. For this, 24 GIFT Nile tilapia from the same spawning stock were divided into two groups based on skin color, namely a black variety (dark skin) and a red variety (reddish skin). There were no significant differences in biometric parameters between varieties. Fish of the red variety had higher 2 h post-slaughter pH and muscle antioxidant capacity and lower yellow-blue axis value (b*), luminosity (L*), thawing loss, cooking loss, and conjugated diene content (P < 0.05). The skin of black tilapia exhibited higher force in the progressive tear test tear, and higher strength, tensile strength, deformation, and elongation in the traction and stretching test than that of red tilapia. Black tilapia skin had higher collagen and hydroxyproline contents. The skin of the red variety exhibited higher expression levels of alpha 1, 2, and 3 type I collagen genes, lower expression levels of the vimentin gene, and higher antioxidant capacity. The results suggest that skin color phenotype may be linked to important metabolic pathways influencing fish fillet and skin traits. These findings can support future research aimed at identifying optimal varieties of Nile tilapia for specific purposes and optimizing the utilization of filleting waste.

An analytical delay model for multi-class and lane-free traffic condition

This study emphasises the criticality of delay as a performance metric for signalized intersections and the challenges associated with its estimation, particularly in the context of Multi-class and Lane-free (MCLF) traffic conditions. Traditional delay models are often inadequate for such conditions, necessitating the development of a tailored approach. A novel delay equation is proposed, integrating insights from queuing theory principles with consideration of multi-class of vehicles and lane-free movement. Key features include assumption of random arrival and departure pattern as well as distribution, incorporation of Passenger Car Equivalent (PCE) and virtual lane concepts to account for the diverse vehicle classes and lane-free movement prevalent in Indian traffic. The model’s efficacy is demonstrated through comparison with conventional in practice delay models, showing its superior performance. This tailored approach enhances the accuracy of delay estimation and also highlights the importance of accounting for specific traffic characteristics in optimising signal design for intersections under MCLF traffic conditions.

Cross-validation of prediction equations for estimating the body mass index in adults without the use of body weight

Introduction Body Mass Index (BMI) is a widely accepted measure by the World Health Organization for assessing body composition, as it provides critical insights into health risks, life expectancy, and quality of life. However, in resource-limited settings, access to weighing scales is often inadequate, and environmental conditions, such as unstable terrain, may hinder accurate weight measurements. In these contexts, alternative methods for estimating BMI become essential for effective health assessment. This study aimed to develop and validate equations to estimate BMI without relying on body weight, providing a practical tool for nutritional assessment where traditional methods are not feasible. Materials and methods Adults aged 18 to 59 of both sexes were included. Variables like waist circumference, height, hip circumference, age, and weight were used for equation development and validation. Participants were divided by sex, with regression and validation subgroups for each. Statistical tests included Student’s t-tests, Pearson correlation, Stepwise Regression, Intraclass Correlation Coefficient, Weighted Kappa Coefficient, and Bland-Altman statistics. Results The study included 810 adults, with 63% (576) women. No significant differences were found in paired comparisons between regression and validation subgroups for both sexes (p > 0.05). Four equations were proposed for BMI estimation: EM2 and EM3 for males, and EF2 and EF3 for females. All equations showed strong positive correlations (r > 0.90), significant at p < 0.05. Regression analysis revealed R2 values between 0.861 and 0.901 (p < 0.000). Intraclass Correlation Coefficient values indicated agreement of 0.961 and 0.972 (p < 0.05), with Weighted Kappa values showing substantial agreement of 0.658 and 0.711 for both sexes (p < 0.05). Conclusion Adopting the proposed equations for estimating BMI in adults without using body weight is safe and effective for measuring this body measure in this population, particularly when weighing these individuals is not feasible.

Decoupled Classifier Knowledge Distillation

Mainstream knowledge distillation methods primarily include self-distillation, offline distillation, online distillation, output-based distillation, and feature-based distillation. While each approach has its respective advantages, they are typically employed independently. Simply combining two distillation methods often leads to redundant information. If the information conveyed by both methods is highly similar, this can result in wasted computational resources and increased complexity. To provide a new perspective on distillation research, we aim to explore a compromise solution that aligns complex features without conflicting with output alignment. In this work, we propose to decouple the classifier’s output into two components: non-target classes learned by the student, and target classes obtained by both the teacher and the student. Finally, we introduce Decoupled Classifier Knowledge Distillation (DCKD), where on one hand, we fix the correct knowledge that the student has already acquired, which is crucial for merging the two methods; on the other hand, we encourage the student to further align its output with that of the teacher. Compared to using a single method, DCKD achieves superior results on both the CIFAR-100 and ImageNet datasets for image classification and object detection tasks, without reducing training efficiency. Moreover, it allows relational-based and feature-based distillation to operate more efficiently and flexibly. This work demonstrates the great potential of integrating distillation methods, and we hope it will inspire future research.

HIV retesting prevalence among clients accessing anti-retroviral therapy and HIV testing services in Ghana

Introduction Ghana is working towards achieving the 95-95-95 targets for its HIV response. One challenge has been low linkage to care rates, possibly due to high rates of retesting among people living with HIV who are already aware of their status. This leads to an overestimation of the first 95 and a subsequent underestimation of the second 95. This study aimed to measure the prevalence of HIV retesting among PLHIV in Ghana who are already aware of their status and to explore their reasons for retesting. Methods This was a facility-based cross-sectional study conducted in the three ecological zones of Ghana. A total of 11,145 individuals from 30 ART centres and 90 HTS centres participated. The sample size for each zone was determined proportionally based on the number of people enrolled in ART. A profiling tool was used to assess testing behaviours among clients visiting HTS sites linked to ART clinics. Focus group discussions were also conducted with clients and health workers to gather their perceptions of reasons for retesting. Results Participants were predominantly female (74.3%; 8,285/11,145), with a median age (interquartile range) of 43.0 (35–52). The prevalence of retesting among ART clients was 32.9% [95% CI: 0.32–0.34] (3,670/11,145). Among those who retested, the majority did so twice (2,041; 55.6%). Of the clients who tested positive for HIV during the study period, 53.1% (43/835; 95% CI: 0.49–0.57) had a previous HIV diagnosis. Adjusting for retesting, the positivity rate at HTS sites decreased from 8.4% to 4.1%. Key reasons for retesting included the desire to confirm diagnosis, denial and doubt regarding test results, retesting required due to documentation issues, and religious beliefs. Conclusion The prevalence of retesting over the past six years was found to be high, resulting in an overestimation of HIV positivity rates and affecting linkage to care. Implementing interventions to accurately account for retesting instances may improve data accuracy and the country’s linkage to care rate, bringing Ghana closer to achieving the 95-95-95 targets.

A single dose of inactivated influenza virus vaccine expressing COBRA hemagglutinin elicits broadly-reactive and long-lasting protection

Influenza virus infections present a pervasive global health concern resulting in millions of hospitalizations and thousands of fatalities annually. To address the influenza antigenic variation, the computationally optimized broadly reactive antigen (COBRA) methodology was used to design influenza hemagglutinin (HA) or neuraminidase (NA) for universal influenza vaccine candidates. In this study, whole inactivated virus (WIV) or split inactivated virus (SIV) vaccine formulations expressing either the H1 COBRA HA or H3 COBRA HA were formulated with or without an adjuvant and tested in ferrets with pre-existing anti-influenza immunity. A single dose of the COBRA-WIV vaccine elicited a robust and broadly reactive antibody response against H1N1 and H3N2 influenza viruses. In contrast, the COBRA-SIV elicited antibodies that recognized fewer viruses, but with R-DOATP, its specificity was expanded. Vaccinated ferrets were protected against morbidity and mortality following challenge with A/California/07/2009 at 14 weeks post-vaccination with reduced viral shedding post-infection compared to the naïve ferrets. However, the COBRA-IIVs did not block the viral transmission to naïve ferrets. The contact infection induced less severe disease and delayed viral shedding than direct infection. Overall, the COBRA HA WIV or the COBRA HA SIV plus R-DOTAP elicited broadly reactive antibodies with long-term protection against viral challenge and reduced viral transmission following a single dose of vaccine in ferrets pre-immune to historical H1N1 and H3N2 influenza viruses. IMPORTANCE The goal of the next-generation influenza vaccine is to provide broadly reactive protection against various drifted influenza strains. With the previous studies evaluating the COBRA HA-based vaccines, the breadth of antibody activities was confirmed following two or three vaccinations. However, for the commercial influenza vaccine, only one shot is required. In this study, only one shot was administrated to the pre-immune ferrets and the COBRA-WIV efficiently elicited broadly reactive antibodies and long-lasting protection against the pdm09 strain. Moreover, this study showed that different infection methods can lead to different disease severity, which emphasizes the significance of the model selection. The infection was conducted 14 weeks post-vaccination to evaluate the long-term protection elicited by only one vaccination. This is the first longevity study describing the immune responses elicited by COBRA-IIVs in ferrets and provides promising results for the potential clinical utilization.

Chaotic behavior, sensitivity analysis and Jacobian elliptic function solution of M-fractional paraxial wave with Kerr law nonlinearity

This study investigates the paraxial approximation of the M-fractional paraxial wave equation with Kerr law nonlinearity. The paraxial wave equation is most important to describe the propagation of waves under the paraxial approximation. This approximation assumes that the wavefronts are nearly parallel to the axis of propagation, allowing for simplifications that make the equation particularly useful in studying beam-like structures such as laser beams and optical solitons. The paraxial wave equation balances linear dispersion and nonlinear effects, capturing the essential dynamics of wave evolution in various media. It plays a crucial role in understanding phenomena like diffraction, focusing, and self-phase modulation in optical fibers. It substantially contributes to our comprehension of the special characteristics of optical soliton solutions and the dynamics of soliton in a variety of optical systems. We create a range of wave structures using the powerful extended Jacobian elliptic function expansion (EJEFE) method, including periodic waves, lump-periodic waves, periodic breather waves, kink-bell waves, kinky-periodic waves, anti-kinky-periodic waves, double-periodic waves, etc. These solutions have applications in wave dynamics in different optical systems and optical fibre. Furthermore, we investigate chaotic phenomena by analyzing the model qualitatively. We analyze phase portraits in detail for a range of parameter values to provide insights into the behavior of the system. We also investigate the sensitivity analysis for diverse parametric values of the perturbated coefficient. We may use various strategies, including time series and 3D and 2D phase patterns, to identify chaotic and quasi-periodic phenomena by providing an external periodic strength. The above discussion of the suggested method demonstrates adaptability and usefulness in resolving a broad spectrum of mathematics and physical difficulties, indicating its potential for generating such optical solutions.

Improving topic modeling performance on social media through semantic relationships within biomedical terminology

Topic modeling utilizes unsupervised machine learning to detect underlying themes within texts and has been deployed routinely to analyze social media for insights into healthcare issues. However, the inherent messiness of social media hinders the full realization of this technique’s potential. As such, we hypothesized that restricting medical concepts in social media texts to specific related semantic types and applying topic modeling to these concepts could be a feasible approach to overcome the challenge of traditional topic modeling for social media texts. Therefore, we developed a semantic-type-based topic modeling pipeline to discover self-reported health-related topics. This pipeline integrated semantic type information and Systematized Medical Nomenclature for Medicine (SNOMED) precoordinated expressions into a traditional topic modeling approach to enhance effectiveness in clustering meaningful, distinct topics. Using social media texts regarding statins for illustration, we evaluated the efficacy of this new approach and validated a newly identified topic using real-world clinical data. Based on expert evaluations, this approach resulted in more novel, distinguishable, and meaningful health-related topics compared to traditional topic modeling. In addition, our electronic health record validation for a newly identified topic in two real-world clinical databases indicated that statin users had a higher prevalence of depression or anxiety compared to matched non-users. Our results indicate that this new topic modeling pipeline can improve the extraction of themes from noisy online discussions, thereby contributing to deeper insights for healthcare research.

Deep vein thrombosis in patients with patellar fractures: Assessing incidence rates and identifying risk factors

Background Deep Vein Thrombosis (DVT) represents a significant complication following orthopedic injuries, particularly patellar fractures. Despite the prevalence, comprehensive studies assessing the incidence rates and identifying specific risk factors in patellar fracture patients are limited. Methods This retrospective analysis reviewed electronic medical records from 3311 patients treated for patellar fractures at two tertiary hospitals between November 2013 and January 2023. The study focused on patient demographics, fracture characteristics, comorbidities, and laboratory parameters to evaluate the incidence and predictors of DVT. DVT prophylaxis measures and diagnostic criteria, including Doppler Ultrasound Scans, were rigorously applied. Results In patients with patellar fractures, the DVT incidence was 30.8%, with 1,790 clots identified in 1,021 diagnosed individuals, predominantly on the injured side (96.7%), and a minor portion on the uninjured side (3.2%). Key risk factors included older age (P<0.001, OR = 1.038), the presence of open injuries (P = 0.002, OR = 1.521), multiple injuries (P<0.001, OR = 3.623), and prolonged time from injury to surgical treatment (P<0.001, OR = 1.097). Conversely, higher levels of albumin (ALB) (P = 0.029, OR = 0.983) and sodium (Na) (P = 0.028, OR = 0.971) were identified as protective factors against DVT. Besides, ROC curve analysis revealed that the age of 52 years and a duration of 4 days from injury to surgery serve as predictive cut-off values for assessing the risk of DVT. Conclusion Our study investigates the incidence of thrombosis in patellar fracture patients and identifies key risk factors for DVT, including age, open and multiple injuries, and the time from injury to surgery. Additionally, we highlight sodium and albumin levels as protective factors. By establishing threshold values for age and surgical delay, our findings improve DVT risk assessment, facilitating earlier and more targeted interventions.

Fingerprinting conflict: A comparative model with applications to archaeological and historical data

This paper is envisioned as a primarily methodological contribution towards a more sophisticated and systematic approach to conflict research in archaeology and history. Studies of conflicts in these fields have often focused on violence and war. Instead, we offer a more holistic approach to conflict research, taking into account different levels of both escalation and de-escalation that embrace all the possible aspects of a conflict from a mere undeveloped potential over complete annihilation to various countermeasures and stages of resolution. A model taking into account different levels of escalation and de-escalation is presented which embodies our multi-faceted view of conflicts and which also allows for a systematic, comparative analysis of conflict situations anywhere and any time in (pre)history. Through ten relatively detailed European case studies spanning the Bronze Age to the 20th century we demonstrate the comparative potential of our model and suggest ways in which it may help to identify typical patterns in conflict situations.

Terahertz Irradiation Promotes Angiogenesis in vitro by Enhancing Permeability of the Voltage-Gated Calcium Channel

Terahertz (THz) waves, positioned between microwave and infrared in the electromagnetic spectrum, have promising applications in medical imaging and biomedicine. In this study, terahertz irradiation at 2.52 THz (100 mW/cm2) did not alter the proliferation of human umbilical vein endothelial cells (HUVECs), but significantly enhanced their angiogenic capacity. This enhancement was accompanied by increased levels of angiogenesis-related proteins such as VEGF in the culture supernatant. ATAC sequencing and RNA sequencing revealed a significant increase in the expression of cytoskeleton-associated genes, including PDXP and SH3BP1, post-irradiation. Additionally, intracellular calcium concentration, closely linked to angiogenesis, markedly increased following terahertz exposure. However, diltiazem significantly mitigated the enhanced angiogenic capacity induced by terahertz irradiation. In conclusion, terahertz irradiation promotes angiogenesis in HUVECs, partly by activating the VEGF signaling pathway through increased calcium fluxes.

Calculating punitive damage multiplier in intellectual property cases: An empirical study and the enhanced model

The application of punitive damages in intellectual property cases in China has encountered notable deficiencies within the judicial context. These deficiencies, including a limited range of applicability, frequent recourse to statutory damages, difficulties in ascertaining the appropriate damages base, and a marked reliance on subjective factors in calculating multipliers, undermine judicial consistency and fairness. In this study, we conducted a comprehensive quantitative analysis by selecting 79 pertinent judgments from a dataset of 3,478 intellectual property rulings. Initially, we developed a multiple linear regression model to assess punitive damages. Upon encountering negative estimated coefficients, we improved the model to a Beta Generalized Linear Model. The efficacy of the new model was validated through an improved R-squared value compared to the original model and by analyzing an additional 45 cases. Both models highlight the scope of infringement and the reputation of intellectual property assets as critical variables. They suggest that wider infringements lead to greater economic penalties and emphasize the significant market value and potential financial losses of intellectual property. Significantly, this study represents a pioneering initiative in China, establishing a model to assist in the legal determination of punitive damages in IP cases.

Expression of concern: Differential effects of UCHL1 modulation on alpha-synuclein in PD-like models of alpha-synucleinopathy

Innovative approaches in QSPR modelling using topological indices for the development of cancer treatments

This paper provides a comprehensive review of quantitative structure-property relationships (QSPR) about to cancer drugs, with a focus on the application of topological indices (TI) and data analysis techniques. Cancer is a serious and life-threatening disease for which no complete cure currently exists. Consequently, extensive research is ongoing to develop new therapeutic agents. The application of topological indices in chemistry and medicine, particularly in the investigation of the molecular, pharmacological, and therapeutic properties of drugs, has become a significant tool. This article investigates the potential of Temperature indices in analyzing the physicochemical properties of drugs used for cancer treatment. The approach employs QSPR modeling to establish correlations between the molecular structure of a compound and its physical and chemical properties. The analysis covers a range of Cancer drugs, including Aminopterin, Convolutamide A, Convolutamydine A, Daunorubicin, Minocycline, Podophyllotoxin, Caulibugulone E, Perfragilin A, Melatonin, Tambjamine K, Amathaspiramide E, and Aspidostomide E. The findings demonstrate that optimal regression models (Fifty-eight models) incorporating TI can effectively predict physicochemical properties, such as Boiling Point (BP), Enthalpy (EN), Flash Point (FP), Molar Refractivity (MR), Polar Surface Area (PSA), Surface Tension (ST), Molecular Volume (MV), and Complexity (COM). This research suggests that temperature-based topological indices (TI) are promising tools for the development and optimization of cancer drugs, as demonstrated by statistically significant results with a p-value less than 0.05. In addition to the linear regression model, which performed the best, two other machine learning models, namely SVR and Random Forest, were also used for further analysis and comparison of their performance in predicting the physicochemical properties of drugs, to assess the advantages and disadvantages of each model.

Centering voices of scientists from marginalized backgrounds to understand experiences in climate adaptation science and inform action

Identifying and building solutions to help people and ecosystems adapt to climate change requires participation of all people; however, Science, Technology, Engineering, and Mathematics (STEM) fields, including environmental sciences, continue to lack diversity. To address this issue, many institutions have increased programming to recruit and retain people from historically marginalized backgrounds in STEM fields. Institutions use surveys to evaluate the experiences of community members and identify areas for improvement; however, surveys often summarize and reflect majority perspectives and disregard voices of historically marginalized individuals. In June 2021, a survey of graduate students, postdocs, faculty, staff, and researchers affiliated with the Northeast Climate Adaptation Science Center (NE CASC) evaluated their experiences of diversity, equity, inclusion, and justice (DEIJ) using Likert-based and long-answer questions. We analyzed the results as a whole, but also focused on the responses of people who self-identified as members of a marginalized group (“marginalized respondents”) in climate adaptation science to center their voices. Marginalized respondents reported being motivated to enter climate adaptation science to improve society and the environment rather than for intellectual curiosity, which motivated one third of non-marginalized respondents. Once in science, marginalized respondents reported feeling less supported and comfortable at work and were more likely to have considered leaving science and academia in the last year. Long-answer responses of marginalized respondents indicated distrust in the ability of leadership and existing DEIJ initiatives to effectively tackle systemic issues and emphasized the importance of focusing on equity and inclusion before recruitment. Marginalized respondents identified additional funding to support existing DEIJ efforts and undergraduates as priorities. By allowing participants to self-identify as part of a marginalized group, we were able to highlight experiences and needs without risking exposure based on race, gender, disability status, or sexual orientation. This approach can be applied to other small organizations with limited demographic diversity.