Knowledge mapping of B cell and atherosclerosis over the past 20 years: A bibliometric analysis.

Atherosclerosis (AS) is the main underlying cause of cardiovascular disease, and B cells are considered a key immune cell type to regulate AS. So far, there is no bibliometric study on B cell and AS. This study aims to comprehensively analyze the scientific output about B cell and AS, summarize the literature characteristics, explore research hotspots, and point out emerging trends. We searched the literature from 2003 to 2022 from the Web of Science Core Collection (WoSCC) database. CiteSpace, VOSviewer, and the R package "Bibliometrix" were used for literature analysis and visualization. A total of 1,062 articles and reviews were identified. The number of annual publications generally showed an upward trend. The United States and China were the most productive countries. Medical University of Vienna was the most productive research institution, and Binder Christoph J. was the most productive author, who was also from Medical University of Vienna. "Arteriosclerosis Thrombosis and Vascular Biology" was the most published journal and the most frequently cited journal. The most cited reference was written by Caligiuri G (2002) in "Journal of Clinical Investigation." The most frequent keywords were "inflammation," "macrophages," "cardiovascular disease," "T cells," "apoptosis," "immunity," "cytokines," "lymphocytes," etc. The trend topics were mainly focused on "immune infiltration," "immunoglobulins," and "biomarkers." The complex role of B cell subtypes and a variety of B cell mediators is the main research direction at present. In-depth analysis of B cell-specific targets can provide new ideas and methods for the prevention and treatment of AS.

A fusion learning method to subgroup analysis of Alzheimer's disease.

Uncovering the heterogeneity in the disease progression of Alzheimer's is a key factor to disease understanding and treatment development, so that interventions can be tailored to target the subgroups that will benefit most from the treatment, which is an important goal of precision medicine. However, in practice, one top methodological challenge hindering the heterogeneity investigation is that the true subgroup membership of each individual is often unknown. In this article, we aim to identify latent subgroups of individuals who share a common disorder progress over time, to predict latent subgroup memberships, and to estimate and infer the heterogeneous trajectories among the subgroups. To achieve these goals, we apply a concave fusion learning method to conduct subgroup analysis for longitudinal trajectories of the Alzheimer's disease data. The heterogeneous trajectories are represented by subject-specific unknown functions which are approximated by B-splines. The concave fusion method can simultaneously estimate the spline coefficients and merge them together for the subjects belonging to the same subgroup to automatically identify subgroups and recover the heterogeneous trajectories. The resulting estimator of the disease trajectory of each subgroup is supported by an asymptotic distribution. It provides a sound theoretical basis for further conducting statistical inference in subgroup analysis.

Mirikizumab Regulates Genes Involved in Ulcerative Colitis Disease Activity and Anti-TNF Resistance: Results From a Phase 2 Study.

INTRODUCTION: Mirikizumab, a monoclonal antibody targeting the p19 subunit of interleukin (IL)-23, demonstrated efficacy and was well-tolerated in a phase 2 randomized clinical trial in patients with moderate-to-severe ulcerative colitis (UC) (NCT02589665). We explored gene expression changes in colonic tissue from study patients and their association with clinical outcomes. METHODS: Patients were randomized to receive intravenous placebo or 3 mirikizumab induction doses. Patient biopsies were collected at baseline and week 12, and differential gene expression was measured using a microarray platform and compared in all treatment groups to determine differential expression values between baseline and week 12. RESULTS: The greatest improvement in clinical outcomes and placebo-adjusted change from baseline in transcripts at week 12 was observed in the 200 mg mirikizumab group. Transcripts significantly modified by mirikizumab correlate with key UC disease activity indices (modified Mayo score, Geboes score, and Robarts Histopathology Index) and include MMP1, MMP3, S100A8, and IL1ß. Changes in transcripts associated with increased disease activity were decreased after 12 weeks of mirikizumab treatment. Mirikizumab treatment affected transcripts associated with resistance to current therapies, including IL-1ß, OSMR, FCGR3A and FCGR3B, and CXCL6, suggesting that anti-IL23p19 therapy modulates biological pathways involved in resistance to antitumor necrosis factor and Janus kinase inhibitors. DISCUSSION: This is the first large-scale gene expression study of inflamed mucosa from patients with UC treated with anti-IL23p19 therapy. These results provide molecular evidence for mucosal healing from an extensive survey of changes in transcripts that improve our understanding of the molecular effects of IL-23p19 inhibition in UC.

Baricitinib for systemic lupus erythematosus: a double-blind, randomised, placebo-controlled, phase 3 trial (SLE-BRAVE-I).

BACKGROUND: Baricitinib is an oral selective inhibitor of Janus kinase 1 and 2 approved for the treatment of rheumatoid arthritis, atopic dermatitis, and alopecia areata. In a 24-week phase 2 study in patients with systemic lupus erythematosus (SLE), baricitinib 4 mg significantly improved SLE disease activity compared with placebo. The objective of this trial was to evaluate the efficacy and safety of baricitinib in patients with active SLE in a 52-week phase 3 study. METHODS: In a multicentre, double-blind, randomised, placebo-controlled, parallel-group, phase 3 study, SLE-BRAVE-I, patients (aged ≥18 years) with active SLE receiving stable background therapy were randomly assigned 1:1:1 to baricitinib 4 mg, 2 mg, or placebo once daily for 52 weeks with standard of care. Glucocorticoid tapering was encouraged but not required per protocol. The primary endpoint was the proportion of patients reaching an SLE Responder Index (SRI)-4 response at week 52 in the baricitinib 4 mg treatment group compared with placebo. The primary endpoint was assessed by logistic regression analysis with baseline disease activity, baseline corticosteroid dose, region, and treatment group in the model. Efficacy analyses were done on a modified intention-to-treat population, comprising all participants who were randomly assigned and received at least one dose of investigational product. Safety analyses were done on all randomly assigned participants who received at least one dose of investigational product and who did not discontinue from the study for the reason of lost to follow-up at the first post-baseline visit. This study is registered with ClinicalTrials.gov, NCT03616912. FINDINGS: 760 participants were randomly assigned and received at least one dose of baricitinib 4 mg (n=252), baricitinib 2 mg (n=255), or placebo (n=253). A significantly greater proportion of participants who received baricitinib 4 mg (142 [57%]; odds ratio 1·57 [95% CI 1·09 to 2·27]; difference with placebo 10·8 [2·0 to 19·6]; p=0·016), but not baricitinib 2 mg (126 [50%]; 1·14 [0·79 to 1·65]; 3·9 [-4·9 to 12·6]; p=0·47), reached SRI-4 response compared with placebo (116 [46%]). There were no significant differences between the proportions of participants in either baricitinib group reaching any of the major secondary endpoints compared with placebo, including glucocorticoid tapering and time to first severe flare. 26 (10%) participants receiving baricitinib 4 mg had serious adverse events, 24 (9%) participants receiving baricitinib 2 mg, and 18 (7%) participants receiving placebo. The safety profile of baricitinib in participants with SLE was consistent with the known baricitinib safety profile. INTERPRETATION: The primary endpoint in this study was met for the 4 mg baricitinib group. However, key secondary endpoints were not. No new safety signals were observed. FUNDING: Eli Lilly and Company.

Baricitinib for systemic lupus erythematosus: a double-blind, randomised, placebo-controlled, phase 3 trial (SLE-BRAVE-II).

BACKGROUND: Baricitinib is an oral selective inhibitor of Janus kinase 1 and 2 approved for the treatment of rheumatoid arthritis, atopic dermatitis, and alopecia areata. In a 24-week phase 2 study in patients with systemic lupus erythematosus (SLE), baricitinib 4 mg significantly improved SLE disease activity compared with placebo. In this Article, we report the evaluation of efficacy and safety of baricitinib in patients with SLE in a 52-week phase 3 study. METHODS: In this phase 3 double-blind, randomised, placebo-controlled study, SLE-BRAVE-II, patients (aged ≥18 years) with active SLE receiving stable background therapy were randomly assigned 1:1:1 to baricitinib 4 mg, baricitinib 2 mg, or placebo once daily for 52 weeks. The primary endpoint was the proportion of patients with an SLE Responder Index (SRI)-4 response at week 52 in the baricitinib 4 mg treatment group compared with placebo. Glucocorticoid tapering was encouraged but not required per protocol. The primary endpoint was assessed by logistic regression analysis with baseline disease activity, baseline corticosteroid dose, region, and treatment group in the model. Efficacy analyses were done on an intention-to-treat population, comprising all participants who were randomly assigned and received at least one dose of investigational product and who did not discontinue from the study for the reason of lost to follow-up at the first post-baseline visit. Safety analyses were done on all randomly assigned participants who received at least one dose of investigational product and who did not discontinue. This study is registered with ClinicalTrials.gov, NCT03616964, and is complete. FINDINGS: A total of 775 patients were randomly assigned and received at least one dose of baricitinib 4 mg (n=258), baricitinib 2 mg (n=261), or placebo (n=256). There was no difference in the primary efficacy outcome of the proportion of SRI-4 responders at week 52 between participants who received baricitinib 4mg (121 [47%]; odds ratio 1·07 [95% CI 0·75 to 1·53]; difference with placebo 1·5 [95% CI -7·1 to 10·2]), 2 mg (120 [46%]; 1·05 [0·73 to 1·50]; 0·8 [-7·9 to 9·4]) and placebo (116 [46%]). None of the major secondary endpoints, including glucocorticoid tapering and time to first severe flare, were met. Serious adverse events were observed in 29 (11%) participants in the baricitinib 4 mg group, 35 (13%) in the baricitinib 2 mg group, and 22 (9%) in the placebo group. The safety profile of baricitinib in patients with SLE was consistent with the known baricitinib safety profile. INTERPRETATION: Although phase 2 data suggested baricitinib as a potential treatment for patients with SLE, which was supported in SLE-BRAVE-I, this result was not replicated in SLE-BRAVE-II. No new safety signals were observed. FUNDING: Eli Lilly and Company.

Patient Disease Trajectories in Rheumatoid Arthritis Patients Treated with Baricitinib 4-mg in Four Phase 3 Clinical Studies.

INTRODUCTION: We sought to identify and compare treatment response groups based on individual patient responses (rather than group mean response) over time on the Clinical Disease Activity Index (CDAI) for rheumatoid arthritis (RA), in patients treated with baricitinib 4-mg in 4 phase 3 studies. METHODS: Trajectory subgroups were identified within each study using growth mixture modeling. Following grouping, baseline characteristics and disease measures were summarized and compared. RESULTS: In each study, three response trajectories were identified. In the three studies of patients naïve to biological disease-modifying anti-rheumatic drugs (bDMARDs) patients had, on average, high disease activity, as measured by CDAI. In these studies, a group of rapid responders (65-71% of patients) had the lowest baseline CDAI scores and achieved mean CDAI ≤ 10 by week 16. Gradual responders (10-17%) had higher baseline CDAI, but generally achieved low disease activity (CDAI ≤ 10) by week 24. A group of partial responders (18-22%) had higher baseline CDAI and did not achieve mean CDAI ≤ 10. In bDMARD-experienced patients, the subgroups were rapid responders, who achieved mean CDAI ≤ 10 (42% of patients); partial responders, with mean CDAI decrease of ~ 15 points from baseline (42% of patients); and limited responders (15% of patients). Changes in modified total sharp score (mTSS; assessed only in biologic-naïve patients) were below the smallest detectable difference at 24/52 weeks for > 90% of patients in each group, excepting partial responders in RA-BEGIN (≥ 75% no detectable change). CONCLUSION: In patients receiving baricitinib 4-mg, lower baseline CDAI was generally associated with rapid response, while higher baseline CDAI scores were generally seen for patients who either reached treatment targets more gradually, or who had a partial or limited response. Maintenance of response was observed with continued baricitinib treatment in all response groups and generally included maintenance of mTSS.

Baricitinib is an oral agent widely approved for the treatment of moderately to severely active rheumatoid arthritis). Although baricitinib (and other agents) have demonstrated efficacy at the population level, treatment responses vary considerably between individual patients. This study assessed four baricitinib phase 3 clinical studies and categorized patient responses into response groups based on the Clinical Disease Activity Index (CDAI) using a growth mixture model. We then evaluated baseline characteristics and corresponding disease measures within the response groups. In patients with no prior treatment with biological disease-modifying anti-rheumatic drugs (bDMARDs), 65­71% of patients had rapid responses to treatment, while smaller groups had gradual (10­17%) or partial (18­22%) responses. In patients with prior bDMARD experience, rapid and partial responders each comprised 42% of patients while 15% had limited response. Gradual responders generally had higher baseline CDAI versus rapid responders, but achieved low disease activity (LDA) by 24, versus 12 weeks for rapid responders. Across response groups, patients who continued treatment generally maintained their response up to 52 weeks, and where joint erosion was assessed (in bDMARD-naïve patients), generally saw maintenance of joints during continued therapy. The identification of a gradual responder group, which demonstrated good response but required more time to achieve LDA, is relatively novel and should be considered when setting treatment expectations, particularly in patients with high baseline disease activity. In addition, in bDMARD-experienced patients, many patients did not achieve LDA but maintained a substantial partial response with continued therapy.

Baricitinib decreases anti-dsDNA in patients with systemic lupus erythematosus: results from a phase II double-blind, randomized, placebo-controlled trial.

BACKGROUND: Patients with systemic lupus erythematosus (SLE) have substantial unmet medical need. Baricitinib is a Janus kinase (JAK)1 and 2 inhibitor that was shown to have therapeutic benefit in patients with SLE in a phase II clinical trial. The purpose of this study was to evaluate the median change from baseline in conventional serologic biomarkers in subgroups and the overall population of baricitinib-treated patients with SLE, and the SLE Responder Index-4 (SRI-4) response by normalization of anti-dsDNA. METHODS: Data were assessed from the phase II trial I4V-MC-JAHH (NCT02708095). The median change from baseline in anti-dsDNA, IgG, and other conventional serologic markers was evaluated over time in patients who had elevated levels of markers at baseline, and in all patients for IgG. Median change from baseline for baricitinib treatments were compared with placebo. Among patients who were anti-dsDNA positive at baseline, SRI-4 responder rate was compared for those who stayed positive or achieved normal levels by week 24. RESULTS: Significant decreases of anti-dsDNA antibodies were observed in response to baricitinib 2 mg and 4 mg compared to placebo beginning at weeks 2 (baricitinib 2 mg = - 14.3 IU/mL, placebo = 0.1 IU/mL) and 4 (baricitinib 4 mg = - 17.9 IU/mL, placebo = 0.02 IU/mL), respectively, continuing through week 24 (baricitinib 2 mg = - 29.6 IU/mL, baricitinib 4 mg = - 15.1 IU/mL, placebo=3.0 IU/mL). Significant reductions from baseline of IgG levels were found for baricitinib 4 mg-treated patients compared to placebo at weeks 12 (baricitinib 4 mg = - 0.65 g/L, placebo = 0.09 g/L) and 24 (baricitinib 4 mg = - 0.60 g/L, placebo = - 0.04 g/L). For patients who were anti-dsDNA positive at baseline, no relationship between achieving SRI-4 responder and normalization of anti-dsDNA was observed by week 24. CONCLUSIONS: Baricitinib treatment resulted in a rapid and sustained significant decrease in anti-dsDNA antibodies compared to placebo among those with positive anti-dsDNA antibodies at baseline, as well as a significant decrease in IgG levels in the 4 mg group at weeks 12 and 24. These data suggest that baricitinib may influence B cell activity in SLE. Further studies are needed to evaluate if reductions in anti-dsDNA levels with baricitinib treatment reflect the impact of baricitinib on B cell activity. TRIAL REGISTRATION: NCT02708095 .

Achieving pain control in early rheumatoid arthritis with baricitinib monotherapy or in combination with methotrexate versus methotrexate monotherapy.

OBJECTIVES: This post hoc analysis assessed speed, magnitude and maintenance of pain improvement in patients with early rheumatoid arthritis (RA) receiving baricitinib, baricitinib and methotrexate (MTX), or MTX over 1 year. Cumulative pain and quality of life benefits were also assessed. METHODS: Randomised, double-blind, phase 3 study RA-BEGIN (NCT01711359) compared baricitinib 4 mg (N=159), baricitinib 4 mg +MTX (N=215) and MTX (N=210) in patients with RA who had no or limited prior disease-modifying antirheumatic drug treatment. Pain was assessed on a 0-100 mm Visual Analogue Scale (VAS). Proportion of patients with ≥30%, ≥50% and ≥70% pain improvement from baseline; ≤20 mm and ≤10 mm on the pain VAS; and time to achieve pain improvement thresholds were assessed over 52 weeks, as were Patient Global Assessment (PtGA) and 36-Item Short Form Health Survey Physical Component Score (SF-36 PCS) outcomes. RESULTS: Baricitinib monotherapy or combination with MTX provides greater (least square mean changes (LSM) from baseline -40 mm and -43 mm, respectively) and more rapid (median 12 and 8 weeks to ≥70% improvement, respectively) pain relief than MTX alone (LSM -31 mm, median 20 weeks to ≥70% improvement) over 52 weeks. Baricitinib, alone or combination, provides 9-10 additional weeks of limited to no pain, similar gain in achievable wellness measured through PtGA, and 5-7 additional weeks with change in SF-36 PCS ≥5 vs MTX over 1 year. CONCLUSIONS: Patients treated with baricitinib reported significantly greater and more rapid pain relief, more weeks with limited to no pain, and clinically meaningful improvements in physical health than patients treated with MTX alone over 1 year.

Maintenance of Patient-Reported Outcomes in Baricitinib-Treated Patients with Moderate-to-Severe Active Rheumatoid Arthritis: Post Hoc Analyses from Two Phase 3 Trials.

INTRODUCTION: Baricitinib has been shown to improve patient-reported outcomes (PROs) in patients with rheumatoid arthritis (RA) who are inadequate responders (IR) to conventional synthetic and biologic disease-modifying antirheumatic drugs (csDMARDs and bDMARDs, respectively). We assessed the ability of baricitinib 2-mg to maintain minimal clinically important differences (MCIDs) in PROs until week 24 among week 4 and 12 responders. METHODS: Data were from two phase 3 trials, RA-BUILD (NCT01721057; csDMARD-IR patients) and RA-BEACON (NCT01721044; bDMARD-IR patients). PROs included Pain Visual Analogue Scale, Health Assessment Questionnaire-Disability Index, Functional Assessment of Chronic Illness Therapy-Fatigue, Short-Form 36 Physical Component Score, and Patient's Global Assessment of Disease Activity. Outcomes were evaluated by proportions of patients achieving MCID improvements, number needed to treat (NNT) at weeks 4, 12, and 24, proportions of patients maintaining MCID responses at week 24 among week 4 or 12 responders, and median time to achieve substantial response with baricitinib 2-mg versus placebo. RESULTS: A higher proportion of baricitinib-treated patients achieved MCID improvements, with NNTs ranging from 5 to 8 for baricitinib 2-mg versus placebo at week 24. Generally, early MCID responses in PROs at weeks 4 or 12 were better maintained through week 24 in RA patients treated with baricitinib 2-mg versus placebo. Patients treated with baricitinib 2-mg also achieved substantial PRO responses or normative values more quickly than placebo. CONCLUSIONS: These results suggest baricitinib-treated patients with RA achieving MCID improvement in PROs at weeks 4 and 12 maintained those improvements over time and that substantial PRO responses were achieved quickly.

Efficacy of Long-Term Treatment with Once-Daily Baricitinib 2 mg in Patients with Active Rheumatoid Arthritis: Post Hoc Analysis of Two 24-Week, Phase III, Randomized, Controlled Studies and One Long-Term Extension Study.

INTRODUCTION: To evaluate long-term efficacy of once-daily baricitinib 2 mg in patients with active rheumatoid arthritis who had an inadequate response (IR) to conventional synthetic disease-modifying antirheumatic drugs (csDMARD) or biologic DMARDs (bDMARD). METHODS: Data from patients treated with baricitinib 2 mg daily in two 24-week, phase III studies, RA-BUILD (csDMARD-IR; NCT01721057) and RA-BEACON (bDMARD-IR; NCT01721044), and one long-term extension study (RA-BEYOND; NCT01885078), were analyzed (120 weeks). The main outcomes were achievement of low-disease activity (LDA; Simple Disease Activity Index [SDAI] ≤ 11), clinical remission (SDAI ≤ 3.3), Health Assessment Questionnaire Disability Index (HAQ-DI) ≤ 0.5 and improvement from baseline of ≥ 0.22, and safety. Analysis populations included (1) all patients and (2) never-rescued patients. Completer and non-responder imputation (NRI) analyses were conducted on each population. RESULTS: In RA-BUILD, 684 were randomized (229 to baricitinib 2 mg, 180 of whom completed RA-BUILD and entered RA-BEYOND). In RA-BEACON, 527 were randomized (174 to baricitinib 2 mg, 117 of whom completed RA-BEACON and entered RA-BEYOND). In RA-BUILD-BEYOND, 85.1% (63/74, completer) and 27.5% (63/229, NRI) of csDMARD-IR patients treated with baricitinib 2 mg achieved SDAI LDA; 40.5% (30/74, completer) and 13.1% (30/229, NRI) were in SDAI remission; 62.2% (46/74, completer) and 20.1% (46/229, NRI) had HAQ-DI ≤ 0.5 and 81.1% (60/74, completer); and 26.2% (60/229, NRI) achieved ≥ 0.22 change from baseline at week 120. In RA-BEACON-BEYOND, 86.5% (32/37, completer) and 18.4% (32/174, NRI) of bDMARD-IR patients treated with baricitinib 2 mg achieved SDAI LDA; 24.3% (9/37, completer) and 5.2% (9/174, NRI) were in SDAI remission; 50.0% (19/38, completer) and 10.9% (19/174, NRI) had HAQ-DI ≤ 0.5; and 73.7% (28/38, completer) and 16.1% (28/174, NRI) achieved ≥ 0.22 change from baseline at week 120. Rates of adverse events of special interest were consistent with previous reports. CONCLUSIONS: Long-term treatment with baricitinib 2 mg demonstrated efficacy for up to 120 weeks and was well tolerated. TRIAL REGISTRATION: ClinicalTrials.gov identifier, NCT01721057, NCT01721044, and NCT01885078.

Fast hybrid Bayesian integrative learning of multiple gene regulatory networks for type 1 diabetes.

Motivated by the study of the molecular mechanism underlying type 1 diabetes with gene expression data collected from both patients and healthy controls at multiple time points, we propose a hybrid Bayesian method for jointly estimating multiple dependent Gaussian graphical models with data observed under distinct conditions, which avoids inversion of high-dimensional covariance matrices and thus can be executed very fast. We prove the consistency of the proposed method under mild conditions. The numerical results indicate the superiority of the proposed method over existing ones in both estimation accuracy and computational efficiency. Extension of the proposed method to joint estimation of multiple mixed graphical models is straightforward.

Efficacy of baricitinib in patients with moderate-to-severe rheumatoid arthritis with 3 years of treatment: results from a long-term study.

OBJECTIVE: To evaluate the long-term efficacy of once-daily baricitinib 4 mg in patients with active RA who were either naïve to DMARDs or who had inadequate response (IR) to MTX. METHODS: Analyses of data from two completed 52-week, phase III studies, RA-BEGIN (DMARD-naïve) and RA-BEAM (MTX-IR), and one ongoing long-term extension (LTE) study (RA-BEYOND) were performed (148 total weeks). At week 52, DMARD-naïve patients treated with MTX monotherapy or baricitinib 4 mg+MTX in RA-BEGIN were switched to open-label baricitinib 4 mg monotherapy; MTX-IR patients treated with adalimumab (+MTX) in RA-BEAM were switched to open-label baricitinib 4 mg (+MTX) in the LTE. Patients who received placebo (+MTX) were switched to baricitinib 4 mg (+MTX) at week 24. Low disease activity (LDA) [Simple Disease Activity Index (SDAI) ≤11], clinical remission (SDAI ≤ 3.3), and physical functioning [Health Assessment Questionnaire Disability Index (HAQ-DI) ≤ 0.5] were assessed. Data were assessed using a non-responder imputation. RESULTS: At week 148, SDAI LDA was achieved in up to 61% of DMARD-naïve patients and 59% of MTX-IR patients initially treated with baricitinib, and SDAI remission was achieved in up to 34% of DMARD-naïve patients and 24% of MTX-IR patients; HAQ-DI ≤ 0.5 was reached in up to 48% of DMARD-naïve patients and 38% of MTX-IR patients initially treated with baricitinib. Over 148 weeks, 3.6% and 10.7% of MTX-IR patients discontinued across treatment groups due to lack of efficacy or due to adverse events, respectively; discontinuation rates were similar in the DMARD-naïve population. CONCLUSION: Treatment with baricitinib 4 mg demonstrated efficacy for up to 3 years and was well tolerated.

Joint estimation of multiple mixed graphical models for pan-cancer network analysis.

Graphical models have been used in many scientific fields for exploration of conditional independence relationships for a large set of random variables. Although a variety of methods have been proposed in the literature for estimating graphical models with different types of data, none of them is applicable for jointly estimating multiple mixed graphical models. To tackle this problem, we propose a joint mixed learning method. The proposed method is very flexible, which works for various mixed types of data, such as those mixed with Gaussian, multinomial, and Poisson, and also allows people to incorporate domain knowledge into network construction by restricting some links to be included in or excluded from the networks. As an application, the proposed method is applied to pan-cancer network analysis for six types of cancer with data from The Cancer Genome Atlas. To our knowledge, this is the first work for joint estimation of multiple mixed graphical models.

Markov Neighborhood Regression for High-Dimensional Inference.

This paper proposes an innovative method for constructing confidence intervals and assessing p-values in statistical inference for high-dimensional linear models. The proposed method has successfully broken the high-dimensional inference problem into a series of low-dimensional inference problems: For each regression coefficient ß i , the confidence interval and p-value are computed by regressing on a subset of variables selected according to the conditional independence relations between the corresponding variable X i and other variables. Since the subset of variables forms a Markov neighborhood of X i in the Markov network formed by all the variables X 1, X 2, … , X p , the proposed method is coined as Markov neighborhood regression. The proposed method is tested on high-dimensional linear, logistic and Cox regression. The numerical results indicate that the proposed method significantly outperforms the existing ones. Based on the Markov neighborhood regression, a method of learning causal structures for high-dimensional linear models is proposed and applied to identification of drug sensitive genes and cancer driver genes. The idea of using conditional independence relations for dimension reduction is general and potentially can be extended to other high-dimensional or big data problems as well.

Joint Bayesian-Incorporating Estimation of Multiple Gaussian Graphical Models to Study Brain Connectivity Development in Adolescence.

Adolescence is a transitional period between the childhood and adulthood with physical changes, as well as increasing emotional development. Studies have shown that the emotional sensitivity is related to a second period of rapid brain growth. However, there is little focus on the trend of brain development during this period. In this paper, we aim to track functional brain connectivity development from late childhood to young adulthood. Mathematically, this problem can be modeled via the estimation of multiple Gaussian graphical models (GGMs). However, most existing methods either require the graph sequence to be fairly long or are only applicable to small graphs. In this paper, we adapted a Bayesian approach incorporating joint estimation of multiple GGMs to overcome the short sequence difficulty, which is also computationally efficient. The data used are the functional magnetic resonance imaging (fMRI) images obtained from the publicly available Philadelphia Neurodevelopmental Cohort (PNC). They include 855 individuals aged 8-22 years who were divided into five different adolescent stages. We summarized the networks with global measurements and applied a hypothesis test across age groups to detect the developmental patterns. Three patterns were detected and defined as consistent development, late puberty, and temporal change. We also discovered several anatomical areas, such as the middle frontal gyrus, putamen gyrus, right lingual gyrus, and right cerebellum crus 2 that are highly involved in the brain functional development. The functional networks, including the salience, subcortical, and auditory networks are significantly developing during the adolescent period.

Learning Moral Graphs in Construction of High-Dimensional Bayesian Networks for Mixed Data.

Bayesian networks have been widely used in many scientific fields for describing the conditional independence relationships for a large set of random variables. This letter proposes a novel algorithm, the so-called p-learning algorithm, for learning moral graphs for high-dimensional Bayesian networks. The moral graph is a Markov network representation of the Bayesian network and also the key to construction of the Bayesian network for constraint-based algorithms. The consistency of the p-learning algorithm is justified under the small-n, large-p scenario. The numerical results indicate that the p-learning algorithm significantly outperforms the existing ones, such as the PC, grow-shrink, incremental association, semi-interleaved hiton, hill-climbing, and max-min hill-climbing. Under the sparsity assumption, the p-learning algorithm has a computational complexity of O(p2) even in the worst case, while the existing algorithms have a computational complexity of O(p3) in the worst case.

An imputation-regularized optimization algorithm for high dimensional missing data problems and beyond.

Missing data are frequently encountered in high dimensional problems, but they are usually difficult to deal with by using standard algorithms, such as the expectation-maximization algorithm and its variants. To tackle this difficulty, some problem-specific algorithms have been developed in the literature, but there still lacks a general algorithm. This work is to fill the gap: we propose a general algorithm for high dimensional missing data problems. The algorithm works by iterating between an imputation step and a regularized optimization step. At the imputation step, the missing data are imputed conditionally on the observed data and the current estimates of parameters and, at the regularized optimization step, a consistent estimate is found via the regularization approach for the minimizer of a Kullback-Leibler divergence defined on the pseudocomplete data. For high dimensional problems, the consistent estimate can be found under sparsity constraints. The consistency of the averaged estimate for the true parameter can be established under quite general conditions. The algorithm is illustrated by using high dimensional Gaussian graphical models, high dimensional variable selection and a random-coefficient model.

Learning gene regulatory networks from next generation sequencing data.

In recent years, next generation sequencing (NGS) has gradually replaced microarray as the major platform in measuring gene expressions. Compared to microarray, NGS has many advantages, such as less noise and higher throughput. However, the discreteness of NGS data also challenges the existing statistical methodology. In particular, there still lacks an appropriate statistical method for reconstructing gene regulatory networks using NGS data in the literature. The existing local Poisson graphical model method is not consistent and can only infer certain local structures of the network. In this article, we propose a random effect model-based transformation to continuize NGS data and then we transform the continuized data to Gaussian via a semiparametric transformation and apply an equivalent partial correlation selection method to reconstruct gene regulatory networks. The proposed method is consistent. The numerical results indicate that the proposed method can lead to much more accurate inference of gene regulatory networks than the local Poisson graphical model and other existing methods. The proposed data-continuized transformation fills the theoretical gap for how to transform discrete data to continuous data and facilitates NGS data analysis. The proposed data-continuized transformation also makes it feasible to integrate different types of data, such as microarray and RNA-seq data, in reconstruction of gene regulatory networks.