Dissertation Defense Announcements

High-quality assembly and annotation of the genome of a species are critical in understanding the genetic basis of almost all aspects of the biology of the species. Although many genome assembly pipelines have been developed, they are either difficult to use or their assemblies are too fragmental. Moreover, although gene annotation pipelines have been developed at large genome centers, they are either too complicated for individual labs to use or not available to public. In this dissertation project, we have proposed a user-friendly pipeline that can assemble genome at chromosome-level with high-quality using PacBio/Nanopore long reads, Illumina paired-end short reads and Hi-C paired-end short reads. We have also developed an accompanying gene annotation pipeline using a combination of homology-based and RNA-based approached. The pipeline achieves high accuracy in protein-coding gene and pseudogene annotations.
Moreover, although multiple chicken genomes have been assembled, high-quality indigenous chicken genomes are still lacking, hampering the understanding of chicken domestication and evolution. Using the pipelines, we assembled and annotated the genomes of four indigenous chickens with distinct morphological traits at the chromosome-level. Our results challenge two earlier conclusions regarding chicken domestication and evolution. First, we found a total of 1,420 new protein-coding genes in the four chickens and recovered 51 of the 274 “missing” genes in birds in general and 36 of the 174 “missing” genes in chickens in particular. Most of these new genes are also found in previously assembled GRCg6a and GRCg7b/w chicken genomes, and might play house-keeping roles. Counting these new genes, chicken genomes encode more genes than originally thought. Second, we identified a total of 2,015 non-processed pseudogenes in the seven genomes. Most pseudogenization mutations are fixed in their respective populations and preferentially occur at the two ends of genes. Purifying selection is relaxed on the pseudogenes, suggesting that they might lose their gene functions. Pseudogenization mutations segregate in the chickens as their phylogenetic tree does, which is based on more than 6,000 essential protein-coding genes. Thus, in contrast to the previous conclusion, loss-of-function mutations play a critical role in chicken domestication and evolution. Moreover, these assembled genomes are valuable resources for studying chicken domestication and evolution.
Furthermore, although many studies related to artificial selection signatures of commercial and indigenous chickens have been carried out, quite a small number of genes have been found to be under selection. To fill these gaps, we re-sequenced 85 individuals of five indigenous chicken breeds with distinct traits from Yunnan, a southwest province of China. By analyzing these indigenous chickens together with 116 individuals of commercial chickens (broilers and layers) and 35 individuals of red jungle fowl (RJF), we find a substantially large number of selective sweeps and affected genes for each chicken breed using a rigorous statistic model than previously reported. We confirm most of previously identified selective sweeps and affected genes. Meanwhile, the vast majority (~98.3%) of our identified selective sweeps overlap known chicken quantitative trait loci. Thus, our predictions are highly reliable. For each breed, we also identify candidate genes and selective sweeps that might be related to the unique traits of the chickens. Most of these genes do not contain nonsense mutations, we therefore quantified the expression levels of eight genes in relevant tissues using RT-qPCR and found that most of them showed differential expression compared to their counterparts.
Finally, eared pheasant species are closely related but inhabit at highly varying altitudes from northeast to southwest China. To understand genetic bases of closely related species to adapt to different altitudes, we sequenced a population of 10 white eared pheasants (WT) (Crossoptilon crossoptilon) inhibiting in 3,000~4,300 m altitude niches in Yunnan, China, and assembled the genome of an individual at chromosome-level with a contig N50 of 19.63 Mb, a scaffold N50 of 29.59 Mb, and a total length of 1.02 Gb. This assembly with only few gaps is of higher quality than a previous one of a brown eared pheasant (BR) (C. mantchuricum) individual living at 20~1000 m altitude in northeast China. Interestingly, the WT genome encodes more protein genes than the BR genome (16,315 VS. 15,003), while the later contains more pseudogenes than the former (1,519 VS. 1,976). The two genomes shared 14,178 genes and 1,040 pseudogenes with 2,137 and 825 unique genes, and 479 and 936 unique pseudogenes, respectively. The unique genes and unique pseudogenes of both species are mainly involved in biological pathways of cardiovascular, energy metabolic, neuronal and immune functions, which are known to be related to adaptation to high altitude. Moreover, we compared the selective sweeps in the genomes of WT, BR and an additional species blue eared pheasant (BL) (C. auritum) inhibiting at 1,500~3,000 m altitude in central west China, using re-sequencing data of 10 WT, 12 BL and 41 BR individuals, respectively. Interestingly, genes under selection in each species converge on the same pathways of the aforementioned four functional categories. These results suggest that these species adapted to highly varying altitudes by loss-of-function mutation and fine-tuning of genes in these common pathways. Our assembled WT genome and re-sequencing data can be valuable resources for studying the biology, evolution and developing conservation strategies of these endangered species.

Nearly 65% of adults report experiencing at least one adverse childhood experience (ACEs). Women are more likely to report experiencing 4 or more ACEs. While the association between ACEs and adverse physical and mental health outcomes in adulthood is well supported, few studies have examined the impact of ACEs on reproductive, prenatal, and perinatal health. Women with a history of ACEs have increased odds of unintended pregnancy, pregnancy complications, and delivering infants who are low birth weight and preterm birth. The purpose of this dissertation was to assess the associations between ACEs and contraceptive use, early initiation of prenatal care (PNC), and delivering a small for gestational age (SGA) infant.
Three separate population-based studies were conducted to investigate these associations using Add Health Public-Use Data, which is a subset of publicly available data from the full National Longitudinal Study of Adolescent Health dataset. The first study assessed the association between ACEs and contraceptive use. The second study evaluated the association between ACEs and early initiation of PNC. The third study examined the ACEs-SGA association as well as examined race/ethnicity as an effect modifier of this association. Logistic regression and multivariate logistic regression were used to calculate the unadjusted and adjusted odds ratios (ORs) and 95% confidence intervals (CIs), respectively. Stratified analysis by race/ethnicity was conducted on the ACEs-SGA association. Given the complex sampling design of Add Health Public-Use Data, all analyses were performed using SAS survey procedures (version 9.4, SAS Institute Inc. Cary, NC).
Across all studies, ACEs were associated with adverse health outcomes. In the first study, women with a family history of suicidal behavior had statistically significant decreased odds of contraceptive use (AOR=0.69, 95% CI: 0.51-0.96). Findings from the second study demonstrated that women who experienced parental alcoholism had 82% statistically significant decreased odds of early initiation of PNC (AOR=0.18, 95% CI: 0.06-0.55). In the third study on ACEs and birth outcomes, women who experienced parental alcoholism had statistically significant increased odds of delivering an SGA infant (AOR=4.11, 95% CI: 1.09-15.52). When stratified by race/ethnicity, among Non-Hispanic White women, those who experienced parental alcoholism had 7-fold statistically significant increased odds of delivering an SGA infant (AOR=7.39, 95% CI: 1.44-37.88). Among Non-Hispanic Black/Hispanic/Other women, those who experienced parental alcoholism had 1.6-fold increased odds of delivering an SGA infant (AOR=1.55, 95% CI: 0.22-10.84).
This dissertation addresses existing gaps in the literature on the impact of ACEs on women’s reproductive, prenatal, and perinatal health. Study results highlight the importance of integrating mental health and reproductive health care services. By implementing trauma-informed care practices such as ACEs screening during reproductive health and PNC visits, healthcare providers may provide additional support for this high-risk population of women. In addition, healthcare providers should underscore the importance of PNC during preconception reproductive health counseling as these visits may serve as an opportunity to engage these women before a pregnancy. By doing so, early PNC may reduce and prevent SGA births.

Social media platforms such as Facebook, TikTok and Instagram have witnessed increasing use of offensive language by online users, which can be harmful to other users. Recently the continuance of the pandemic has propelled the propagation of offensive content associated with Covid-19 on social media. Some researchers begin to develop effective methods for detecting online offensive language from social media content automatically, yet automatic intervention of offensive language after it is detected remains largely understudied. To address the gaps, this dissertation develops an effective text style transfer-based approach, TST-IOC, for automatic offensive intervention tasks. The promising outcome suggests that our proposed method shows significant potential and could be a preferred choice among users for offensive intervention tasks. This dissertation provides some contributions. First, it contributes significantly to the field of offensive language research by introducing a novel text style transfer-based approach, which has been rarely explored in existing intervention studies. This approach shows a step forward in the development of an automatic offensive intervention system, addressing the limitations of current filtering systems deployed by social media platforms. Second, existing research has mainly focused on using performance metrics for evaluating offensive intervention methods quantitatively. However, this study goes beyond by proposing a comprehensive automatic evaluation paradigm. By exploring both quantitative and qualitative aspects of automatic intervention assessment, it fills a crucial gap in the current offensive language research landscape. Finally, it recognizes the scarcity of studies comparing human evaluation with automatic evaluation in automatic intervention systems. To bridge this gap, we conduct a user study, which allows for an investigation of user acceptance of the proposed automatic intervention approach in real-world scenarios. The insights gained from this user study not only guide the design of more comprehensive automatic intervention systems but also hold the potential to shape the development of human-centric automatic intervention systems in the future.

Additive manufacturing, specifically the Fused Deposition Modeling (FDM) method, has emerged as a promising technique for manufacturing. Using FDM, complex geometries can be created using precise layer-by-layer deposition of material. Among the advantages of this method are its cost-effectiveness, rapid prototyping capabilities, and ability to customize. Due to the similar melting point of ferroelectric polymers PVDF and PVDF-TrFE, which can be used for FDM printers, this study examined the possibility of using FDM for additive manufacturing of PVDF and PVDF-TrFE sensors with enhanced piezoelectric and pyroelectric properties. The resulting sensors can find applications in diverse fields such as biomedical engineering, robotics, energy harvesting, and sensing technologies, enabling advancements in various sectors that require sensitive and reliable sensor systems. Although both PVDF and PVDF-TrFE can be printed by FDM, the XRD result indicated that only PVDF-TrFE crystallized in the polar phase upon cooling from the melt while PVDF always crystallized in the nonpolar phase. Therefore, only PVDF-TrFE could be used for piezoelectric and pyroelectric samples. Using the corona discharge method, consistent responses from both piezoand pyroelectric sensors were observed. Using a 30 mW laser, samples were measured for pyroelectricity. Upon poling at 25 kV for 10 minutes at room temperature, the maximum pyroelectric response was 50 mV. Samples were clamped in one end and measured in deflection mode for their piezoelectric response. Upon stimulating the free end of a PVDF-TrFE sample printed on a PVDF layer as a substrate, 130 V of open circuit piezoelectric response was observed.