Yuliang Shi

Yuliang Shi

Auto-generated profile, based on publicly available information

Yuliang Shi

Department of Statistics & Actuarial Science University of Waterloo Waterloo Canada

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 March 2024 in Blood Advances
Verified
Journal Article
Open access
PDF file
MDMarta B. Davidson
Marta B. Davidson
JKJames A. Kennedy
James A. Kennedy
JCJose-Mario Capo-Chichi
Jose-Mario Capo-Chichi
YSYuliang Shi
Yuliang Shi
WXWei Xu
Wei Xu
VGVikas Gupta
Vikas Gupta
Transformation of BCR::ABL1-negative myeloproliferative neoplasms (MPN) to an accelerated or blast phase is associated with poor outcomes. The efficacy of acute myeloid leukemia (AML)–type intensive and nonintensive hypomethylating agent–based regimens is not well studied. We therefore performed a retrospective analysis of patients with MPN-AP/BP (N = 138) treated with intensive (N = 81) and nonintensive (N = 57) blast-reduction strategies. We used clinically relatable response criteria developed at the Princess Margaret Cancer Centre. The overall best response, comprising complete remission (CR), complete remission with incomplete hematologic recovery (CRi), and reversion to chronic phase MPN (cMPN), in the intensive and nonintensive groups was 77% (62 of 81) and 39% (21 of 54), respectively. Similar overall best response rates were observed in patients receiving induction with daunorubicin combined with cytarabine arabinoside (daunorubicin + ara-C) (74% [23 of 31]) or FLAG-IDA/NOVE-HiDAC (78% [39 of 50], P = .78). However, patients receiving daunorubicin + ara-C more often required second inductions (29% [9 of 31] vs 4% [2 of 50], P = .002). Most responses in the entire cohort were reversions to cMPN (55 of 83 [66%]). CR and CRi comprised 30% (25 of 83) and 4% (3 of 83) of responses, respectively. Mutations in TP53 (overall response [OR] 8.2 [95% confidence interval [CI] 2.01, 37.1], P = .004) and RAS pathway (OR 5.1 [95%CI 1.2, 23.7], P = .03) were associated with inferior treatment response for intensively treated patients, and poorer performance status (Eastern Cooperative Oncology Group) was associated with inferior treatment response in both intensively (OR 10.4 [95% CI 2.0, 78.5], P = .009) and nonintensively treated groups (OR 12 [95% CI 2.04, 230.3], P = .02). In patients with paired samples before and after therapy (N = 26), there was a significant residual mutation burden remaining irrespective of response to blast-reduction therapy.

19 June 2025 in Statistics in Biosciences
Verified
Journal Article
Closed access
PDF file
YSYuliang Shi
Yuliang Shi
YZYEYING ZHU
YEYING ZHU
JDJoel A. Dubin
Joel A. Dubin

21 October 2026 in Canadian Journal of Statistics
Verified
Journal Article
Open access
PDF file
YSYuliang Shi
Yuliang Shi
YZYEYING ZHU
YEYING ZHU
JDJoel A. Dubin
Joel A. Dubin
In causal inference, properly selecting the propensity score (PS) model is an important topic and has been widely investigated in observational studies. However, there are very few studies investigating the model selection issue for causal inference when the exposure is missing at random (MAR). When the exposure variable contains missing data, an intuitive approach is to impute the missing exposure values first and perform a PS analysis on the imputed dataset under MAR assumption. In this article, we discuss how to select both imputation and PS models. We then propose a new criterion, termed the “rank score,” for evaluating the overall performance of both models. Simulation studies show that the full imputation plus the outcome‐related PS models leads to the smallest root mean squared error, and the rank score can help select the best models. An application study is conducted to quantify the causal effect of cardiovascular disease on the mortality of COVID‐19 patients.