Jichun Xie

Most single-cell RNA sequencing (scRNA-seq) analyses begin with cell clustering; thus, the clustering accuracy considerably impacts the validity of downstream analyses. In contrast with the abundance of clustering methods, the tools to assess the clustering accuracy are limited. We propose a new Clustering Deviation Index (CDI) that measures the deviation of any clustering label set from the observed single-cell data. We conduct in silico and experimental scRNA-seq studies to show that CDI can select the optimal clustering label set. As a result, CDI also informs the optimal tuning parameters for any given clustering method and the correct number of cluster components.

We have found a mistake in the proof Theorem 6 in our published paper “Optimal False Discovery Rate Control for Dependent Data” [4].We apologize to the readers and thank Professor Jens Ledet Jensen at Aarhus University for his question which led to identification of this mistake. We provide here a corrected proof of Theorem 6 with further clarifications of the assumptions.

Although hematopoietic stem cell transplantation (HCT) is the only curative treatment for acute myeloid leukemia (AML), it is associated with significant treatment related morbidity and mortality. There is great need for predictive biomarkers associated with overall survival (OS) and clinical outcomes. We hypothesized that circulating metabolic, inflammatory, and immune molecules have potential as predictive biomarkers for AML patients who receive HCT treatment. This retrospective study was designed with an exploratory approach to comprehensively characterize immune, inflammatory, and metabolomic biomarkers. We identified patients with AML who underwent HCT and had existing baseline plasma samples. Using those samples (n = 34), we studied 65 blood based metabolomic and 61 immune/inflammatory related biomarkers, comparing patients with either long-term OS (≥ 3 years) or short-term OS (OS ≤ 1 years). We also compared the immune/inflammatory response and metabolomic biomarkers in younger vs. older AML patients (≤30 years vs. ≥ 55 years old). In addition, the biomarker profiles were analyzed for their association with clinical outcomes, namely OS, chronic graft versus host disease (cGVHD), acute graft versus host disease (aGVHD), infection and relapse. Several baseline biomarkers were elevated in older versus younger patients, and baseline levels were lower for three markers (IL13, SAA, CRP) in patients with OS ≥ 3 years. We also identified immune/inflammatory response markers associated with aGVHD (IL-9, Eotaxin-3), cGVHD (Flt-1), infection (D-dimer), or relapse (IL-17D, bFGF, Eotaxin-3). Evaluation of metabolic markers demonstrated higher baseline levels of medium- and long-chain acylcarnitines (AC) in older patients, association with aGVHD (lactate, long-chain AC), and cGVHD (medium-chain AC). These differentially expressed profiles merit further evaluation as predictive biomarkers.

The APOBEC family of cytidine deaminases is one of the most common endogenous sources of mutations in human cancer. Genomic studies of tumors have found that APOBEC mutational signatures are enriched in the HER2 subtype of breast cancer and are associated with immunotherapy response in diverse cancer types. However, the direct consequences of APOBEC mutagenesis on the tumor immune microenvironment have not been thoroughly investigated. To address this, we developed syngeneic murine mammary tumor models with inducible expression of APOBEC3B. We found that APOBEC activity induced antitumor adaptive immune responses and CD4+ T cell-mediated, antigen-specific tumor growth inhibition. Although polyclonal APOBEC tumors had a moderate growth defect, clonal APOBEC tumors were almost completely rejected, suggesting that APOBEC-mediated genetic heterogeneity limits antitumor adaptive immune responses. Consistent with the observed immune infiltration in APOBEC tumors, APOBEC activity sensitized HER2-driven breast tumors to anti-CTLA-4 checkpoint inhibition and led to a complete response to combination anti-CTLA-4 and anti-HER2 therapy. In human breast cancers, the relationship between APOBEC mutagenesis and immunogenicity varied by breast cancer subtype and the frequency of subclonal mutations. This work provides a mechanistic basis for the sensitivity of APOBEC tumors to checkpoint inhibitors and suggests a rationale for using APOBEC mutational signatures and clonality as biomarkers predicting immunotherapy response in HER2-positive (HER2+) breast cancers.

In many applications, a large number of features are collected with the goal to identify a few important ones. Sometimes, these features lie in a metric space with a known distance matrix, which partially reflects their co-importance pattern. Proper use of the distance matrix will boost the power of identifying important features. Hence, we develop a new multiple testing framework named the Distance Assisted Recursive Testing (DART). DART has two stages. In stage 1, we transform the distance matrix into an aggregation tree, where each node represents a set of features. In stage 2, based on the aggregation tree, we set up dynamic node hypotheses and perform multiple testing on the tree. All rejections are mapped back to the features. Under mild assumptions, the false discovery proportion of DART converges to the desired level in high probability converging to one. We illustrate by theory and simulations that DART has superior performance under various models compared to the existing methods. We applied DART to a clinical trial in the allogeneic stem cell transplantation study to identify the gut microbiota whose abundance will be impacted by the after-transplant care.