Hasil Pencarian  ::  Simpan CSV :: Kembali

"[ABSTRAKBerdasarkan data World Health Organization (WHO) diketahui prevalensi herpesdi negara-negara berkembang lebih tinggi dibandingkan dengan di negara maju.Virus herpes dapat ditemukan dimana saja dan salah satu ciri penting adalahkemampuannya yang dapat menimbulkan infeksi akut dan kronik pada waktuwaktutertentu. Akibat infeksi tersebut memungkinkan terjadi komplikasi yanglebih berat. Virus herpes terdiri atas genome DNA tertutup inti yang mengandungprotein dan dibungkus oleh glikoprotein. Dengan mempelajari ekspresi gen(sekuen DNA/protein) dan didukung oleh kemajuan di bidang bioinformatika,dapat ditemukan sub-sub bagian penting dan kelompok gen. Virus-virus ini dapatdikelompokkan dengan menganalisa sekuens protein dari virus herpes denganmenggunakan algoritma Tribe Markov Clustering (Tribe-MCL). Tribe-MCLmerupakan metode clustering efisien berdasarkan teori rantai Markov chain,untuk mengelompokkan barisan keluarga protein. Data sekuens protein virusherpes diperoleh di GenBank yang dapat diakses pada situs National Center forBiotechnology Information (NCBI), kemudian disejajarkan menggunakanprogram BLASTp. Hasil pengelompokan sekuen protein virus herpesmenggunakan algoritma Tribe-MCL dengan program R diperoleh enamkelompok . Semua kelompok menunjukkan jenis protein yang sama, dalam halini jenis protein yang digunakan adalah glikoprotein B, M, dan H pada delapanjenis virus herpes yang terjangkit pada manusia.

---

ABSTRACTBased on World Health Organization (WHO) data, the prevalence of herpes indeveloping countries is higher than in developed countries. The herpes virus canbe found anywhere and one of the important characteristics is its ability to causeacute and chronic infection at certain times. Due to infections enables moresevere complications occur. The herpes virus is composed of DNA containingprotein and wrapped by glycoproteins. By studying the expression of genes(sequences of DNA / protein) and is supported by advances in bioinformatics, canbe found an important sub-sections and groups of genes. These viruses can beclassified by analyzing the sequence of the protein-sequence of the herpes virususing algorithm Tribe Markov Clustering (Tribe-MCL). Tribe-MCL is an efficientclustering method based on the theory of Markov chains, to classify sequences ofprotein families. Herpes virus protein sequence data obtained in GenBank whichcan be accessed on the website National Center for Biotechnology Information(NCBI), then aligned using BLASTp program. The results of clustering proteinsequences herpes virus using algorithms (Tribe-MCL) with a program of Robtained six cluster. All clusters showed the same type of protein, in this case thetype of protein used is a glycoprotein B, F, and H in eight types of herpes virusthat infected humans, Based on World Health Organization (WHO) data, the prevalence of herpes indeveloping countries is higher than in developed countries. The herpes virus canbe found anywhere and one of the important characteristics is its ability to causeacute and chronic infection at certain times. Due to infections enables moresevere complications occur. The herpes virus is composed of DNA containingprotein and wrapped by glycoproteins. By studying the expression of genes(sequences of DNA / protein) and is supported by advances in bioinformatics, canbe found an important sub-sections and groups of genes. These viruses can beclassified by analyzing the sequence of the protein-sequence of the herpes virususing algorithm Tribe Markov Clustering (Tribe-MCL). Tribe-MCL is an efficientclustering method based on the theory of Markov chains, to classify sequences ofprotein families. Herpes virus protein sequence data obtained in GenBank whichcan be accessed on the website National Center for Biotechnology Information(NCBI), then aligned using BLASTp program. The results of clustering proteinsequences herpes virus using algorithms (Tribe-MCL) with a program of Robtained six cluster. All clusters showed the same type of protein, in this case thetype of protein used is a glycoprotein B, F, and H in eight types of herpes virusthat infected humans]"

"Systemic Lupus Erythematosus (SLE) merupakan penyakit autoimun kronis di mana sistem imun secara keliru menyerang sel-sel sehat dalam tubuh. Pada tahun 2017, survei yang dilakukan oleh Profesor Handono Kalim dan rekan-rekannya memperkirakan bahwa prevalensi SLE di Indonesia mencapai 0,5% dari total populasi, dengan jumlah penderita diperkirakan lebih dari 1,3 juta orang. Meskipun demikian, karena SLE merupakan penyakit autoimun reumatik, pengobatan SLE tidak ditujukan untuk penyembuhan total, melainkan untuk meredakan gejala, mencegah kerusakan organ, meningkatkan kualitas hidup pasien, serta mencapai kondisi remisi.  Pada penyakit SLE, interaksi protein-protein memiliki peran penting dalam memahami jalur patogenesis. Oleh karena itu, deteksi interaksi protein–protein memegang peranan penting dalam mengidentifikasi target-target terapeutik potensial serta mendukung pengembangan terapi yang disesuaikan dengan gejala yang dialami oleh masing-masing pasien. Penelitian ini bertujuan untuk mengimplementasikan metode Protein-protein Interaction Prediction Based on Siamese Residual Recurrent Convolutional Neural Network (PIPR), yaitu sebuah model deep learning yang menggunakan arsitektur siamese berbasis residual recurrent convolutional neural network (residual RCNN), untuk mendeteksi kelas interaksi protein-protein yang berhubungan dengan penyakit SLE, sekaligus mengevaluasi performa dari model tersebut. Sebanyak 177 gen manusia yang terkait dengan SLE diidentifikasi melalui basis data Online Mendelian Inheritance in Man (OMIM), dan diperoleh 5311 pasangan interaksi protein dari database Search Tool for the Retrieval of Interacting Genes/Proteins (STRING), yang terdiri atas 4059 pasangan kelas interaksi dan 1252 pasangan kelas non-interaksi. Untuk mengatasi ketidakseimbangan kelas, dilakukan proses balancing data dengan metode NearMiss-2 berdasarkan nilai fitur hydropathy index. PIPR bekerja dengan beberapa tahap. Pertama, setiap ID protein, yaitu kode unik yang diberikan pada protein, digunakan untuk memperoleh sekuens protein dari masing-masing protein. Selanjutnya, sekuens protein tersebut diubah menjadi representasi vektor embedding menggunakan metode seq2tensor, dan diproses oleh model PIPR. Evaluasi kinerja model dilakukan menggunakan metrik akurasi, spesifisitas, sensitivitas, F1-score dan AUC-ROC. Penelitian ini menunjukkan performa terbaik dengan akurasi sebesar 0,9860, sensitivitas 0,9881, spesifisitas 0,9879, F1-score 0,9861, serta AUC-ROC 0,9948. Secara keseluruhan, hasil tersebut menunjukkan bahwa model mampu mendeteksi interaksi protein-protein pada penyakit SLE dengan sangat baik. Nilai sensitivitas yang lebih tinggi dibandingkan spesifisitas mengindikasikan bahwa model memiliki kemampuan yang kuat dalam mengenali pasangan protein yang saling berinteraksi.

---

Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease in which the immune system erroneously attacks the body's healthy cells. In 2017, a survey conducted by Professor Handono Kalim and his colleagues estimated that the prevalence of SLE in Indonesia reached 0,5% of the total population, with the number of affected individuals estimated at more than 1.3 million. However, as a rheumatic autoimmune disease, treatment for SLE is not aimed at achieving full recovery but rather focuses on alleviating symptoms, preventing organ damage, improving patients’ quality of life, and achieving remission. In the context of SLE, protein–protein interactions play an important role in elucidating pathogenic pathways. Therefore, analyzing these interactions holds the potential to provide novel insights for the development of more effective therapies. This study aims to implement the Protei-protein Interaction Prediction based on Siamese Residual Recurrent Convolutional Neural Network (PIPR) method, a deep learning model that uses a siamese architecture with a residual recurrent convolutional neural network (residual RCNN) encoder to detect protein–protein interaction classes relevant to SLE and to evaluate the model’s performance. A total of 177 human genes associated with SLE were identified through the Online Mendelian Inheritance in Man (OMIM) database, yielding 5311 protein interaction pairs from the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database, consisting of 4059 interacting and 1252 non-interacting pairs. To address class imbalance, a data balancing procedure was carried out using the NearMiss-2 method based on hydropathy index values. The PIPR model operates in several stages. First, each protein ID, as a unique protein identifier, is used to retrieve its corresponding sequence. Subsequently, the protein sequences are converted into embedding vector representations using the seq2tensor method and processed through the PIPR model. The model's performance was evaluated using accuracy, specificity, sensitivity, F1-score, and AUC-ROC as the evaluation metrics. This study achieved the best performance with an accuracy of 0,9860, sensitivity of 0,9881, specificity of 0,9879, F1-score of 0,9861, and AUC-ROC of 0,9948. Overall, these results demonstrate that the model is highly effective in detecting protein–protein interactions related to SLE. The higher sensitivity compared to specificity indicates that the model has a strong ability to recognize interacting protein pairs. "