پديد آورنده :
ذاكري، پويا
عنوان :
بهبود پيش بيني مكان استقرار پروتئين در ميتوكندري با استفاده از تركيب داده بر مبناي ويژگي هاي گوناگون توالي هاي پروتئيني
مقطع تحصيلي :
كارشناسي ارشد
گرايش تحصيلي :
هوش مصنوعي و رباتيك
محل تحصيل :
اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر
صفحه شمار :
يازده،209،[II]ص.: مصور،جدول،نمودار
يادداشت :
ص.ع.به فارسي و انگليسي
استاد راهنما :
محمد حسين سراجي، بهزاد مشيري
توصيفگر ها :
ماشين بردارپشتيبان , عملگر ميانگين مرتب وزن دار , همتراز دوبدو
تاريخ نمايه سازي :
3/8/88
استاد داور :
رسول موسوي، نفيسه نيلي
دانشكده :
مهندسي برق و كامپيوتر
چكيده فارسي :
به فارسي و انگليسي: قابل رويت در نسخه ديجيتالي
چكيده انگليسي :
Improvement of protein submitochondria locations prediction using data fusion based on various features of sequence Pooya Zakery p zakery@ec iut ac ir Date of Submission April 2009 Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan 84156 83111 Iran Degree M sc Language FarsiSupervisor Mohammad H Saraee saraee@cc iut ac ir Behzad Moshiri moshiri@ut ac irAbstractOne of the fundamental goals in cell biology and proteomics is to identify the subcellularlocations Information of the subcellular locations for proteins can provide useful cluesabout their functions Another area of which has received little attention in the literature isthe submitochondria locations Mitochondria are subcellular organelles that appear only ineukaryotic cells The mitochondrion is the power generator playing a critical role ingenerating energy in the eukaryotic cell Mitochondria are surrounded by two layers ofmembrane the inner membrane and the outer membrane Proteins which are localizedwithin mitochondria play important roles in energy metabolism process Inner membrane outer membrane and matrix contain proteins which do contribute to different procedures inenergy metabolism The use of a reliable automatic submitochondria localizer could speedup the drugs design for over 100 kinds of complex diseases related to mitochondria likeprogrammed cell death and ionic homeostasis In this thesis we construct predictors for protein submitochondria locations based onvarious features of sequence We have used 11 representation models of protein samplesthat include amino acid composition dipeptide composition higher order dipeptidecomposition compositions of amino acid properties Chou pseudo amino acidcomposition functional domain composition model based on prediction of solventaccessibility model based on prediction of secondary structure elements the combinationdiscrete model based on prediction of solvent accessibility and secondary structureelements discrete model of pairwise sequence alignment and improved discrete model ofpairwise sequence alignment Each biological feature is selected as input to multiclassSupport Vector Machine SVM classifier By using leave one out cross validation theprediction accuracy based on the improved discrete model of pairwise sequence alignmentis 1 better than the best computational system that used for this problem In addition we have selected Ordered Weighted Averaging OWA which is one of fusiondata operators to apply on the 11 best SVM based classifiers which we have constructed inprevious phase The results of 94 01 on prediction accuracy show that our approach issuperior to the result of 89 of the best existing approach reported This improvement ismade possible by using leave one out cross validation Key Words Mitochondria Protein Localization Support Vector Machine Data Fusion OrderedWeighted Averaging Pairwise Alignment
استاد راهنما :
محمد حسين سراجي، بهزاد مشيري
استاد داور :
رسول موسوي، نفيسه نيلي
