빅데이터 활용 및 사업을 위한 비식별화 전략

김정범(Kim JeongBeom),임채원(Lim ChaeWon),하재현(Ha JaeHyub),김문기(Kim MoonKi),박연지(Park YeonJi),신진슬(Shin JinSeol),김유지(Kim Yooji),이단비(Lee Danbi),이진형(Lee Jinhyung),하송미(Ha Songmi),김지현(Kim Jihyon),김은석(Kim Enseok) 한국정보기술학회 2019 Proceedings of KIIT Conference Vol.2019 No.6

빅데이터를 사용할 때 가장 중요한 요소 중 하나는 비식별 전략입니다. 개인 정보를 식별 할 수없는 식별불가능한 정보는 빅데이터 분석 및 출력으로 만 사용할 수 있습니다. 비식별 조치는 대용량 데이터 수집을 위한 개인 정보 적용 정책에 따라 적절하게 수행되어야합니다. 비식별전략은 큰 데이터 세트에서 개인을 식별할 수있는 요소 전부 또는 일부의 삭제, 대체 등을 통해 개인을 식별하는 것을 불가능하게 하는 정책입니다. 비식별 정보는 전략 수립을 통해 개인 정보 이외의 정보로 추정되므로 정보 주체의 동의없이 제 3 자에게 사용 또는 제공 할 수 있습니다. 따라서 빅데이터 분석과 결과의 활용에 있어서 가장 중요한 전략입니다. 식별되지 않은 결과는 비 개인 정보로 간주되지만 새로운 바인딩 기술이 나타나거나 결합 될 수있는 정보가 다시 식별 될 수 있으므로, 필수적인 관리 및 기술 안전장치를 구현해야합니다. One of the most important factors in using big data is the de-identification strategy. Non-identifiable information that does not identify personal information can only be used as a big data analysis and output. De-identification measures should be appropriately performed in accordance with the personal information application policy for the collection of big data types. De-identification is a policy that makes it impossible to identify an individual through deletion, substitution, etc., of all or some of the elements that can identify an individual in a big data set. Since the de-identification information is estimated as information other than personal information by establishing strategy, the information can be used or provided to a third party without consent from the information subject. Therefore, in the analysis of big data and utilization of the result it is the most important strategy. Although non-identified outputs are assumed to be non-personal information, essential management and technical safeguards should be implemented, as new binding techniques may appear or information that can be combined may be re-identified. Through this paper, we will examine the related strategies and implementation example.

The nature of triple-negative breast cancer classification and antitumoral strategies

Kim, Songmi,Kim, Dong Hee,Lee, Wooseok,Lee, Yong-Moon,Choi, Song-Yi,Han, Kyudong Korea Genome Organization 2020 Genomics & informatics Vol.18 No.4

Identifying the patterns of gene expression in breast cancers is essential to understanding their pathophysiology and developing anticancer drugs. Breast cancer is a heterogeneous disease with different subtypes determined by distinct biological features. Luminal breast cancer is characterized by a relatively high expression of estrogen receptor (ER) and progesterone receptor (PR) genes, which are expressed in breast luminal cells. In ~25% of invasive breast cancers, human epidermal growth factor receptor 2 (HER2) is overexpressed; these cancers are categorized as the HER2 type. Triple-negative breast cancer (TNBC), in which the cancer cells do not express ER/PR or HER2, shows highly aggressive clinical outcomes. TNBC can be further classified into specific subtypes according to genomic mutations and cancer immunogenicity. Herein, we discuss the brief history of TNBC classification and its implications for promising treatments.

The Conformity Effect in Online Product Rating: The Pattern Recognition Approach

Kim, Hyung Jun,Kim, Songmi,Kim, Wonjoon The Korea Contents Association 2017 International Journal of Contents Vol.13 No.4

Since the advent of the Internet, and the development of smart devices, people have begun to spend more time in online platforms; this phenomenon has created a large number of online Words of Mouth (WOM) daily. Under these changes, one of the important aspects to consider is the conformity effect in online WOM; that is, whether an individual's own opinion would be influenced by the majority opinion of other people. This study, therefore, investigates whether there is the conformity effect in online product ratings for Amazon.com using the method called Markov Chain analysis. Markov Chain analysis considers the stochastic process that satisfies the Markov property, and we assume that the generation of online product ratings follows the process. Under the assumption that people are usually independent when they express their opinion in online platforms, we analyze the interdependency among rating sequences, and we find weak evidence that there exists the conformity effect in online product rating. This suggests that people who leave online product ratings consider others' opinions.

Re-irradiation of unresectable recurrent head and neck cancer: using Helical Tomotherapy as image-guided intensity-modulated radiotherapy

Songmi Jeong,Eun Jung Yoo,Ji Yoon Kim,Chi Wha Han,Ki Jun Kim,Chul Seung Kay 대한방사선종양학회 2013 Radiation Oncology Journal Vol.31 No.4

Purpose: Re-irradiation (re-RT) is considered a treatment option for inoperable locoregionally recurrent head and neck cancer (HNC) after prior radiotherapy. We evaluated the efficacy and safety of re-RT using Helical Tomotherapy as image-guided intensity-modulated radiotherapy in recurrent HNC. Materials and Methods: Patients diagnosed with recurrent HNC and received re-RT were retrospectively reviewed. Primary endpoint was overall survival (OS) and secondary endpoints were locoregional control and toxicities. Results: The median follow-up period of total 9 patients was 18.7 months (range, 4.1 to 76 months) and that of 3 alive patients was 49 months (range, 47 to 76 months). Median dose of first radiotherapy and re-RT was 64.8 and 47.5 Gy₁₀. Median cumulative dose of the two courses of radiotherapy was 116.3 Gy₁₀ (range, 91.8 to 128.9 Gy₁₀) while the median interval between the two courses of radiation was 25 months (range, 4 to 137 months). The response rate after re-RT of the evaluated 8 patients was 75% (complete response, 4; partial response, 2). Median locoregional relapse-free survival after re-RT was 11.9 months (range, 3.4 to 75.1 months) and 5 patients eventually presented with treatment failure (in-field failure, 2; in- and out-field failure, 2; out-field failure, 1). Median OS of the 8 patients was 20.3 months (range, 4.1 to 75.1 months). One- and two-year OS rates were 62.5% and 50%, respectively. Grade 3 leucopenia developed in one patient as acute toxicity, and grade 2 osteonecrosis and trismus as chronic toxicity in another patient. Conclusion: Re-RT using Helical Tomotherapy for previously irradiated patients with unresectable locoregionally recurrent HNC may be a feasible treatment option with long-term survival and acceptable toxicities.

Investigation of high correlation with carcass traits of SNPs of the PLCB1, C/EBPα, and TDRKH genes and the combinations of SNPs using the MDR method in the Hanwoo

Kim Songmi,Lee Yong-Moon,Kim Dong Hee,Ha Jae Jung,이준구,김대현,Oh Dongyep,한규동 한국유전학회 2021 Genes & Genomics Vol.43 No.8

Background Recently, many researchers focus on the best way to produce high-quality meat, as the trend in food consumption today is to focus on quality. In general, consumers’ preferences in beef difer depending on taste and meatiness. Therefore, researchers are interested in how the marbling score afects the favors of meat or the various factors that make up the meatiness to captivate the consumers’ tastes. Objective This study identifes single nucleotide polymorphisms (SNPs) or gene combinations that afect the carcass traits of Korean cattle (Hanwoo) by using the multifactor dimensionality reduction (MDR) method. Methods We collected the candidate SNPs to identify SNPs related to marbling scores from whole-exome sequencing and bovine SNP genotyping data. Using 96 Hanwoo samples, we performed PCR amplifcation to investigate the polymorphism status. In addition, we investigated genetic relationships between carcass traits and SNPs using 612 Hanwoo samples. Furthermore, each candidate SNP genotype and the combinations of SNP genotypes were verifed to improve the accuracy of genetic relationships using MDR method. Results Twenty-four candidate SNPs associated with carcass trait and marbling scores were identifed from SNP genotyping and whole-exome sequencing. Among them, three SNP markers (c.459 T>C of the PLCB1 gene, c.271 A>C of the C/EBPα gene, and g.17257 A>G of the TDRKH gene) were showed statistically signifcant diferences between intramuscular fat and genotypes. Especially, two candidate SNPs, including c.459 T>C located in the PLCB1 gene and c.271 A>C located in the C/ EBPα gene, could be highly associated with the intramuscular fat of Hanwoo quality grade. In addition, the combination of SNP genotypes is showed higher signifcant diferences with carcass weight, backfat thickness, and longissimus dorsi muscle area. Conclusion Three SNP genotypes and the combination of SNP genotypes in the PLCB1, C/EBPα, and TDRKH genes may be useful genetic markers for improving beef quality.

Comparison of COVID-19-associated multisystem inflammatory syndrome in children (MIS-C) and Kawasaki disease shock syndrome: case reports and literature review

( Songmi Lee ),( Danbi Kim ),( Beom Joon Kim ),( Jung Woo Rhim ),( Soo-young Lee ),( Dae Chul Jeong ) 대한류마티스학회 2023 대한류마티스학회지 Vol.30 No.4

Multisystem inflammatory syndrome in children (MIS-C) is a serious post-infectious complication of COVID-19 characterized by hyperinflammation and multi-organ dysfunction including shock. Shock is also seen in a severe form of Kawasaki disease (KD) called KD shock syndrome (KDSS). Here, we present one MIS-C and one KDSS case and compare similarities and differences between them. Both MIS-C (case 1) and KDSS (case 2) showed hyperinflammation, KD-related features, gastrointestinal problems, hypotension, and coagulopathy. The extent of systemic inflammation and organ dysfunction was more severe in KDSS than in MIS-C. Case 1 was diagnosed as MIS-C because SARS-CoV-2 was confirmed, and case 2 was diagnosed as KDSS because no pathogen was identified in microbiological studies. We believe that the most important difference between MIS-C and KDSS was whether SARS-CoV-2 was identified as an infectious trigger. Organ dysfunction is a hallmark of MIS-C and KDSS, but not KD, so MIS-C shares more clinical phenotypes with KDSS than with KD. Comparison of MIS-C and KDSS will be an interesting and important topic in the field of KD-like hyperinflammatory disease research.

A Novel Competitive Class of α-Glucosidase Inhibitors: (E)-1-Phenyl-3-(4-Styrylphenyl)Urea Derivatives

Kim, Jun Young,Lee, Ji Won,Kim, Young Soo,Lee, Yuno,Ryu, Young Bae,Kim, Songmi,Ryu, Hyung Won,Curtis-Long, Marcus J.,Lee, Keun Woo,Lee, Woo Song,Park, Ki Hun WILEY-VCH Verlag 2010 Chembiochem Vol.11 No.15

<P>Competitive glycosidase inhibitors are generally sugar mimics that are costly and tedious to obtain because they require challenging and elongated chemical synthesis, which must be stereo- and regiocontrolled. Here, we show that readily accessible achiral (E)-1-phenyl-3-(4-strylphenyl)ureas are potent competitive α-glucosidase inhibitors. A systematic synthesis study shows that the 1-phenyl moiety on the urea is critical for ensuring competitive inhibition, and substituents on both terminal phenyl groups contribute to inhibition potency. The most potent inhibitor, compound 12 (IC<SUB>50</SUB>=8.4 μM, K<SUB>i</SUB>=3.2 μM), manifested a simple slow-binding inhibition profile for α-glucosidase with the kinetic parameters k<SUB>3</SUB>=0.005256 μM<SUP>−1</SUP> min<SUP>−1</SUP>, k<SUB>4</SUB>=0.003024 min<SUP>−1</SUP>, and <TEX>${K{{{\rm app}\hfill \atop {\rm i}\hfill}}}$</TEX>=0.5753 μM.</P> <B>Graphic Abstract</B> <P>Friendly competition: Competitive α-glucosidase inhibitors are generally sugar mimics that are costly and tedious to obtain. We have developed a novel competitive class of α-glucosidase inhibitors, stilbene urea derivatives, that manifest a slow-binding inhibition profile (see graph). <img src='wiley_img_2010/14394227-2010-11-15-CBIC201000376-content.gif' alt='wiley_img_2010/14394227-2010-11-15-CBIC201000376-content'> </P>

A comprehensive analysis of chimpanzee (Pan Troglodytes)-specific AluYb8 element

Songmi Kim,Dong Hee Kim,Hiroo Imai,Yong‑Moon Lee,Kyudong Han 한국유전학회 2020 Genes & Genomics Vol.42 No.10

Background Alu elements are most abundant retrotransposons with>1.2 million copies in the primate genome. AluYb8 subfamily was diverged from AluY lineage, and has accumulated eight diagnostic mutations and 7-bp duplication during primate evolution. A total of 1851 AluYb copies are present in the human genome, and most of them are human-specifc. On the other hand, only a few AluYb8 copies were identifed in the chimpanzee genome by previous studies on AluYb8. The signifcantly diferent number of species-specifc AluYb8 elements between human and chimpanzee might result from the incompletion of chimpanzee reference genome sequences at the time of the previous study. Objective AluYb8 elements could generate genomic structural variations in the chimpanzee genome. This study aimed to identify and characterize chimpanzee-specifc AluYb elements using the most updated chimpanzee reference genome sequences (Jan. 2018, panTro6). Methods To identify chimpanzee-specifc AluYb8, we carried out genomic comparison with non-chimpanzee primate genome using the UCSC table browser. In addition, chimpanzee-specifc AluYb8 candidates were manually inspected and experimentally verifed using PCR and Sanger sequencing. Results Among a total of 231 chimpanzee-specifc AluYb8 candidates, 11 of the candidates are chimpanzee-specifc AluYb8, and 29 elements are shared between the chimpanzee and non-chimpanzee primate genomes. Through the sequence analysis of AluYb8 and other Alu subfamilies, we were able to observe various diagnostic mutations and variable length duplications in 7-bp duplication region of AluYb8 element. In addition, we further validated two of the chimpanzee-specifc AluYb8 elements (CS8 and CS20) that were not previously discovered by display PCR and Sanger sequencing. Interestingly, we identifed a AluYb8 insertion-mediated deletion (CS8 locus) in the chimpanzee genome. Conclusion Our study found that AluYb8 elements are much more abundant in the human genome than chimpanzee genome, and that it could be due to the absence of hyperactive “master” AluYb8 elements in the chimpanzee genome.

Prognostic Significance of ARID1A Expression Patterns Varies with Molecular Subtype in Advanced Gastric Cancer

Kim Jun Yong,Park Cheol Keun,Noh Songmi,Cheong Jae-Ho,Noh Sung Hoon,Kim Hyunki 거트앤리버 소화기연관학회협의회 2023 Gut and Liver Vol.17 No.5

Background/Aims: AT-rich interactive domain 1A (ARID1A) is frequently mutated in gastric cancer (GC), especially Epstein-Barr virus (EBV)-associated and microsatellite instability high GC. The loss of ARID1A expression has been reported as a poor prognostic marker in GC. However, the relationships between ARID1A alteration and EBV-associated and microsatellite instability high GC, which are known to have a favorable prognosis, has hampered proper evaluation of the prognostic significance of ARID1A expression in GC. We aimed to analyze the true prognostic significance of ARID1A expression by correcting confounding variables. Methods: We evaluated the ARID1A expression in a large series (n=1,032) of advanced GC and analyzed the relationships between expression pattern and variable parameters, including clinicopathologic factors, key molecular features such as EBV-positivity, mismatch repair protein deficiency, and expression of p53 and several receptor tyrosine kinases including human epidermal growth factor receptor 2, epidermal growth factor receptor, and mesenchymal-epithelial transition factor. Survival analysis of the molecular subtypes was done according to the ARID1A expression patterns. Results: Loss of ARID1A expression was found in 52.5% (53/101) of mutL homolog 1 (MLH1)- deficient and 35.8% (24/67) of EBV-positive GCs, compared with only 9.6% (82/864) of the MLH1-proficient and EBV-negative group (p<0.001). The loss of ARID1A expression was associated only with MLH1 deficiency and EBV positivity. On survival analysis, the loss of ARID1A expression was associated with worse prognosis only in MLH1-proficient and EBV-negative GC. Multivariate analysis revealed that both loss of ARID1A and decreased ARID1A expression were independent worse prognostic factors in patients with advanced GC. Conclusions: Only in MLH1-proficient and EBV-negative GC, the loss of ARID1A expression is related to poorer prognosis.

Structural Variation of Alu Element and Human Disease

Kim, Songmi,Cho, Chun-Sung,Han, Kyudong,Lee, Jungnam Korea Genome Organization 2016 Genomics & informatics Vol.14 No.3

Transposable elements are one of major sources to cause genomic instability through various mechanisms including de novo insertion, insertion-mediated genomic deletion, and recombination-associated genomic deletion. Among them is Alu element which is the most abundant element, composing ~10% of the human genome. The element emerged in the primate genome 65 million years ago and has since propagated successfully in the human and non-human primate genomes. Alu element is a non-autonomous retrotransposon and therefore retrotransposed using L1-enzyme machinery. The 'master gene' model has been generally accepted to explain Alu element amplification in primate genomes. According to the model, different subfamilies of Alu elements are created by mutations on the master gene and most Alu elements are amplified from the hyperactive master genes. Alu element is frequently involved in genomic rearrangements in the human genome due to its abundance and sequence identity between them. The genomic rearrangements caused by Alu elements could lead to genetic disorders such as hereditary disease, blood disorder, and neurological disorder. In fact, Alu elements are associated with approximately 0.1% of human genetic disorders. The first part of this review discusses mechanisms of Alu amplification and diversity among different Alu subfamilies. The second part discusses the particular role of Alu elements in generating genomic rearrangements as well as human genetic disorders.