BackgroundVariation in seed oil composition and content among soybean varieties is largely attributed to differences in transcript sequences and/or transcript accumulation of oil production related genes in seeds. Discovery and analysis of sequence and expression variations in these genes will accelerate soybean oil quality improvement.ResultsIn an effort to identify these variations, we sequenced the transcriptomes of soybean seeds from nine lines varying in oil composition and/or total oil content. Our results showed that 69,338 distinct transcripts from 32,885 annotated genes were expressed in seeds. A total of 8,037 transcript expression polymorphisms and 50,485 transcript sequence polymorphisms (48,792 SNPs and 1,693 small Indels) were identified among the lines. Effects of the transcript polymorphisms on their encoded protein sequences and functions were predicted. The studies also provided independent evidence that the lack of FAD2-1A gene activity and a non-synonymous SNP in the coding sequence of FAB2C caused elevated oleic acid and stearic acid levels in soybean lines M23 and FAM94-41, respectively.ConclusionsAs a proof-of-concept, we developed an integrated RNA-seq and bioinformatics approach to identify and functionally annotate transcript polymorphisms, and demonstrated its high effectiveness for discovery of genetic and transcript variations that result in altered oil quality traits. The collection of transcript polymorphisms coupled with their predicted functional effects will be a valuable asset for further discovery of genes, gene variants, and functional markers to improve soybean oil quality.

Various algorithms in reinforcement learning exhibit dramatic variability in their convergence rates and ultimate accuracy as a function of the problem structure. Such instance-specific behavior is not captured by existing global minimax bounds, which are worst-case in nature. We analyze the problem of estimating optimal Q-value functions for a discounted Markov decision process with discrete states and actions and identify an instance-dependent functional that controls the difficulty of estimation in the 8 -norm. Using a local minimax framework, we show that this functional arises in lower bounds on the accuracy on any estimation procedure. In the other direction, we establish the sharpness of our lower bounds, up to factors logarithmic in the state and action spaces, by analyzing a variance-reduced version of Q-learning. Our theory provides a precise way of distinguishing "easy" problems from "hard" ones in the context of Q-learning, as illustrated by an ensemble with a continuum of difficulty.

Melanoma is commonly driven by activating mutations in the MAP kinase BRAF; however, oncogenic BRAF alone is insufficient to promote melanomagenesis. Instead, its expression induces a transient proliferative burst that ultimately ceases with the development of benign nevi comprised of growth-arrested melanocytes. The tumor suppressive mechanisms that restrain nevus melanocyte proliferation remain poorly understood. Here we utilize cell and murine models to demonstrate that oncogenic BRAF leads to activation of the Hippo tumor suppressor pathway, both in melanocytes in vitro and nevus melanocytes in vivo. Mechanistically, we show that oncogenic BRAF promotes both ERK-dependent alterations in the actin cytoskeleton and whole-genome doubling events, which independently reduce RhoA activity to promote Hippo activation. We also demonstrate that functional impairment of the Hippo pathway enables oncogenic BRAF-expressing melanocytes to bypass nevus formation and rapidly form melanomas. Our data reveal that the Hippo pathway enforces the stable arrest of nevus melanocytes and represents a critical barrier to melanoma development.

ImportanceAlthough isotretinoin may rarely be associated with laboratory abnormalities such as hypertriglyceridemia, the optimal approach to laboratory monitoring is uncertain, and there is wide variation in clinical practice.ObjectiveTo establish a consensus for isotretinoin laboratory monitoring among a diverse, international cohort of clinical and research experts in acne.Design, Setting, and ParticipantsUsing a modified electronic Delphi process, 4 rounds of anonymous electronic surveys were administered from 2021 to 2022. For laboratory tests reaching consensus (≥70% agreement) for inclusion, questions regarding more time-specific monitoring throughout isotretinoin therapy were asked in subsequent rounds. The participants were international board-certified dermatologist acne experts who were selected on a voluntary basis based on involvement in acne-related professional organizations and research.Main Outcomes and MeasuresThe primary outcome measured was whether participants could reach consensus on key isotretinoin laboratory monitoring parameters.ResultsThe 22 participants from 5 continents had a mean (SD) time in practice of 23.7 (11.6) years and represented a variety of practice settings. Throughout the 4-round study, participation rates ranged from 90% to 100%. Consensus was achieved for the following: check alanine aminotransferase within a month prior to initiation (89.5%) and at peak dose (89.5%) but not monthly (76.2%) or after treatment completion (73.7%); check triglycerides within a month prior to initiation (89.5%) and at peak dose (78.9%) but not monthly (84.2%) or after treatment completion (73.7%); do not check complete blood cell count or basic metabolic panel parameters at any point during isotretinoin treatment (all >70%); do not check gamma-glutamyl transferase (78.9%), bilirubin (81.0%), albumin (72.7%), total protein (72.7%), low-density lipoprotein (73.7%), high-density lipoprotein (73.7%), or C-reactive protein (77.3%).Conclusions and RelevanceThis Delphi study identified a core set of laboratory tests that should be evaluated prior to and during treatment with isotretinoin. These results provide valuable data to guide clinical practice and clinical guideline development to optimize laboratory monitoring in patients treated with isotretinoin.

Introduction: To evaluate the use of inaccurate terminology used by dermatology practices to describe the training and qualifications of their nonphysician clinicians (NPCs) when new patients are booking appointments.Methods: Clinics were randomly selected and called to determine the first available appointment for a new patient with a new and changing mole. If the receptionist confirmed the first-offered appointment was with an NPC, the encounter was included in this study. If receptionists used inaccurate terminology to describe the NPCs and their qualifications, this instance was recorded along with the specific language that they used.Results: A total of 344 unique dermatology clinics were contacted on February 27, 2020, in 25 states. Phone calls at 128 clinics (37.2%) met our inclusion criterion. Inaccurate language was used to describe NPCs at 23 (18%) unique clinic locations across 12 states, with "dermatologist," "doctor," "physician," and "boardcertified" being used to describe NPCs as the most common inaccurate terms.Conclusion: These findings demonstrate that front office staff at dermatology clinics use inaccurate and potentially misleading terminology to refer to NPCs working in their clinics. While we cannot establish whether this is intentional or due to a lack of training, additional focus should be placed on accurately representing provider qualifications to patients.

Various algorithms for reinforcement learning (RL) exhibit dramatic variation in their convergence rates as a function of problem structure. Such problem-dependent behavior is not captured by worst-case analyses and has accordingly inspired a growing effort in obtaining instance-dependent guarantees and deriving instance-optimal algorithms for RL problems. This research has been carried out, however, primarily within the confines of theory, providing guarantees that explain ex post the performance differences observed. A natural next step is to convert these theoretical guarantees into guidelines that are useful in practice. We address the problem of obtaining sharp instance-dependent confidence regions for the policy evaluation problem and the optimal value estimation problem of an MDP, given access to an instance-optimal algorithm. As a consequence, we propose a data-dependent stopping rule for instance-optimal algorithms. The proposed stopping rule adapts to the instance-specific difficulty of the problem and allows for early termination for problems with favorable structure.

Background In cosmetic dermatology, lasers and lights treat a variety of hair and skin conditions, including some that disproportionately affect people of color. Aims Our systematic review aims to understand the representation of participants with skin phototypes 4–6 in cosmetic dermatologic trials studying laser and light devices. Methods A systematic literature search was conducted using search terms “laser,” “light,” and multiple laser and light subtypes in the PubMed and Web of Science databases. All randomized controlled trials (RCTs) published between January 1, 2010 and October 14, 2021 that studied laser or light devices for cosmetic dermatologic conditions were eligible for inclusion. Results Our systematic review included 461 RCTs representing 14 763 participants. Of 345 studies that reported skin phototype, 81.7% (n = 282) included participants of skin phototypes 4–6, but only 27.5% (n = 95) included participants of skin phototypes 5 or 6. This trend of excluding darker skin phototypes persisted when results were stratified by condition, laser of study, study location, journal type, and funding source. Conclusions Trials studying lasers and lights for the treatment of cosmetic dermatologic conditions need better representation of skin phototypes 5 and 6.

No abstract