?url_ver=Z39.88-2004&rft_id=1917051024&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=KLASIFIKASI+IMBALANCED+DATA+UNTUK+POST-TRANSLATIONAL%0D%0AMODIFICATION+(PTM)+PADA+SEQUENCES+PROTEIN+LISIN%0D%0AMENGGUNAKAN+ADAPTIVE+SYNTHETIC+(ADASYN)+SAMPLING%0D%0ADENGAN+METODE+LIGHT+GRADIENT+BOOSTING+MACHINE%0D%0A(LightGBM)&rft.creator=Ardella+Dean%2C+Awalia&rft.subject=000+Ilmu+komputer%2C+informasi+dan+pekerjaan+umum&rft.subject=500+ilmu+pengetahuan+alam+dan+matematika&rft.description=Crotonylation+merupakan+salah+satu+jenis+modifikasi+pascatranslasi+(PTM)+berupa%0D%0Apenambahan+gugus+asil+pada+asam+amino+lisin.+Modifikasi+ini+memiliki+kemampuan%0D%0Adalam+mengatur+ekspresi+gen+dan+ditemukan+terlibat+pada+beberapa+penyakit%2C%0D%0Aseperti+depresi%2C+ginjal+kronis%2C+hingga+kanker.+Identifikasi+situs+PTM+menjadi+hal%0D%0Akrusial+mengingat+perannya+dalam+siklus+sel.+Metode+machine+learning+untuk%0D%0Aklasifikasi+situs+PTM+dapat+digunakan+sebagai+alternatif+dalam+mengenali+situs%0D%0APTM%2C+namun+membutuhkan+data+yang+relatif+banyak+agar+dapat+memberikan+hasil%0D%0Ayang+andal.+Penelitian+ini+dilakukan+menggunakan+metode+klasifikasi+LightGBM%0D%0Adan+teknik+oversampling+ADASYN+untuk+menangani+ketersediaan+data+protein%0D%0Acrotonylation+yang+cukup+sedikit.+Data+yang+digunakan+diperoleh+dari+situs+UniProt%0D%0Aterdiri+dari+159+data+positif+dan+847+data+negatif.+Ekstraksi+fitur+menggunakan%0D%0Abinary+encoding%2C+position+weight+amino+acid%2C+encoding+based+on+grouped+weight%2C%0D%0Ak-nearest+neighbors%2C+dan+pseudo-position+specific+scoring+matrix+menghasilkan+833%0D%0Afitur.+5-fold+cross-validation+digunakan+pada+proses+training+untuk+mencari%0D%0Akombinasi+hyperparameter+terbaik.+Hasil+penelitian+menunjukkan+bahwa%0D%0Apembagian+data+menggunakan+90%25+data+sebagai+data+latih+dan+10%25+data+sebagai%0D%0Adata+uji+memberikan+hasil+tertinggi+dengan+nilai+accuracy+sebesar+96%2C04%25%2C%0D%0Asensitivity+sebesar+87%2C50%25%2C+specificity+sebesar+97%2C65%25%2C+MCC+sebesar+85%2C15%25%2C+dan%0D%0AAUC+sebesar+98%2C90%25.%0D%0AKata+kunci%3A+modifikasi+pascatranslasi%2C+crotonylation%2C+ADASYN%2C+LightGBM%0D%0A%0D%0ACrotonylation+is+a+type+of+post-translational+modification+(PTM).+It+is+an+addition%0D%0Aof+acyl+group+to+the+lysine+residues.+This+modification+has+the+ability+to+regulate%0D%0Agene+expression+and+has+been+found+to+be+involved+in+several+diseases%2C+such+as%0D%0Adepression%2C+chronic+kidney+disease%2C+and+cancer.+Identification+of+PTM+sites+is%0D%0Acrucial+considering+their+role+in+the+cell+cycle.+Machine+learning+methods+for%0D%0Aclassifying+PTM+sites+can+be+used+as+an+alternative+for+recognizing+PTM+sites%2C+but%0D%0Athey+require+relatively+large+amounts+of+data+to+provide+reliable+results.+This%0D%0Aresearch+was+carried+out+using+the+LightGBM+classification+method+and+the%0D%0AADASYN+oversampling+technique+to+handle+the+limited+availability+of+crotonylated%0D%0Aprotein+sequences.+The+data+was+obtained+from+the+UniProt+website+consisting+of%0D%0A159+positive+data+(crotonylated)+and+847+negative+data+(noncrotonylated).+Feature%0D%0Aextraction+by+using+binary+encoding%2C+position+weight+amino+acid%2C+encoding+based%0D%0Aon+grouped+weight%2C+k-nearest+neighbors%2C+and+pseudo-position+specific+scoring%0D%0Amatrix+produced+833+features.+5-fold+cross+validation+was+used+in+the+training%0D%0Aprocess+to+find+best+hyperparameter+combinations.+The+results+showed+that+the%0D%0Ahighest+result+was+obtained+by+using+90%25+of+the+data+as+training+data+with+the%0D%0Aapplication+of+ADASYN+oversampling+(n_neighbors%3D9)+and+10%25+of+the+data+as+test%0D%0Adata+with+96%2C04%25+accuracy%2C+87%2C50%25+sensitivity%2C+97.65%25+specificity%2C+85%2C15%25+MCC%2C%0D%0Aand+98%2C90%25+AUC.%0D%0AKeywords%3A+post-translational+modification%2C+crotonylation%2C+ADASYN%2C+LightGBM&rft.publisher=FAKULTAS+MATEMATIKA+DAN+ILMU+PENGETAHUAN+ALAM&rft.date=2023-11-23&rft.type=Skripsi&rft.type=NonPeerReviewed&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F77446%2F1%2FABSTRAK.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F77446%2F2%2FSKRIPSI%2520FULL.pdf&rft.format=text&rft.identifier=http%3A%2F%2Fdigilib.unila.ac.id%2F77446%2F3%2FSKRIPSI%2520TANPA%2520BAB%2520PEMBAHASAN.pdf&rft.identifier=++Ardella+Dean%2C+Awalia++(2023)+KLASIFIKASI+IMBALANCED+DATA+UNTUK+POST-TRANSLATIONAL+MODIFICATION+(PTM)+PADA+SEQUENCES+PROTEIN+LISIN+MENGGUNAKAN+ADAPTIVE+SYNTHETIC+(ADASYN)+SAMPLING+DENGAN+METODE+LIGHT+GRADIENT+BOOSTING+MACHINE+(LightGBM).++FAKULTAS+MATEMATIKA+DAN+ILMU+PENGETAHUAN+ALAM%2C+UNIVERSITAS+LAMPUNG.+++++&rft.relation=http%3A%2F%2Fdigilib.unila.ac.id%2F77446%2F