Search or add a thesis

Advanced Search (Beta)
Home > Prediction of Membrane Proteins Using Machine Learning Approaches

Prediction of Membrane Proteins Using Machine Learning Approaches

Thesis Info

Access Option

External Link

Author

Hayat, Maqsood

Program

PhD

Institute

Pakistan Institute of Engineering and Applied Sciences

City

Islamabad

Province

Pakistan

Country

Pakistan

Thesis Completing Year

2012

Thesis Completion Status

Completed

Subject

Computer Science

Language

English

Link

http://prr.hec.gov.pk/jspui/handle/123456789/1640

Added

2021-02-17 19:49:13

Modified

2024-03-24 20:25:49

ARI ID

1676727813231

Asian Research Index Whatsapp Chanel
Asian Research Index Whatsapp Chanel

Join our Whatsapp Channel to get regular updates.

Similar


Membrane proteins are the basic constituent of a cell that manage intra and extracellular processes of a cell. About 20-30% of genes of eukaryotic organisms are encoded from membrane proteins. In addition, almost 50% of drugs are directly targeted against membrane proteins. Owing to the significant role of membrane proteins in living organisms, the identification of membrane proteins with substantial accuracy is essential. However, the annotation of membrane proteins through conventional methods is difficult, sometimes even impossible. Therefore, membrane proteins are predicted from topogenic sequences using computational intelligence techniques. In this study, we conducted our research in two phases regarding the prediction of membrane protein types and structures. In Phase-I, regarding the prediction of membrane protein types, four different ways are explored in order to enhance true prediction. In the first part of phase-I, membrane protein types are predicted using Composite protein sequence representation followed by the application of principal component analysis in conjunction with individual classifiers. In the second part, the notion of ensemble classification is utilized. In part three, an error correction code is incorporated with Support Vector Machine using evolutionary profiles (Position Specific Scoring Matrix) and SAAC based features. Finally, in part four, a two-layer web predictor Mem- PHybrid is developed. Mem-PHybrid accomplishes the prediction in two steps. First, a protein query is identified as a membrane or a non-membrane protein. In case of membrane protein, then its type is predicted. In the second phase of this research, the structure of membrane protein is recognized as alpha-helix transmembrane or outer membrane proteins. In case of alpha- helix transmembrane proteins, features are explored from protein sequences by two feature extraction schemes of distinct natures; including physicochemical properties and compositional index of amino acids. Singular value decomposition is employed to extract high variation features. A hybrid feature vector is formed by combining the different types of features. Weighted Random Forest is then used as a classification algorithm. On the other hand, in case of outer membrane proteins, protein sequences are represented by Amino acid composition, PseAA composition, and SAAC along with their hybrid models. Genetic programming, K-nearest neighbor, and fuzzy K-nearest neighbor are adopted as classification algorithms. Through the simulation study, we observed that the prediction performance of our proposed approaches in case of both types and structures prediction is better compared to existing state of the arts/approaches. Finally, we conclude that our proposed approach for membrane proteins might play a significant role in Computational Biology, Molecular Biology, Bioinformatics, and thus might help in applications related to drug discovery. In addition, the related web predictors provide sufficient information to researchers and academicians in future research.
Loading...
Loading...

Similar Books

Loading...

Similar Chapters

Loading...

Similar News

Loading...

Similar Articles

Loading...

Similar Article Headings

Loading...

عرضِ مصنف

بندہ ناچیز انتہائی ادب سے عرض گزار ہے کہ اللہ رب العزت نے ہمیں انسان بنایا اور اپنے پیارے محبوب کے دامن رحمت سے وابستہ فرمایا ۔ علاوہ ازیں بے شمار نعمتوں سے نوازا۔
ہمیں چاہیے کہ ہم اپنے رب کی عبادت کریں اور اس کے شکر گزار بندے بنیں۔ شریعت مطہرہ کے مطابق اپنی زندگی بسر کریں، لیکن ہم دین سے بہت دور ہوتے جارہے ہیں ۔ نفس کی خواہشات اور رسم و رواج کو بہت زیادہ اہمیت دیتے ہیںاور دین کے معاملہ بھی اپنی مرضی کرتے ہیں ۔ کسی نے کیا اچھا لکھا ہے:
جیویں پیارا راضی ہووے مرضی ویکھ سجن دی
جئے کر مرضی اپنی لوڑیں ایہہ گل کدے نہ بندی
جب کہ خدمت، ہمدردی ، درد دل کا جذبہ ہونا چاہیے ۔ علماء کا کام تو صرف بتادینا ہے عمل ہم نے کرنا ہے: ’’جیسا کریں گے ویسا بھریں گے‘‘ ۔ ذکر آتا ہے کہ بھگت کبیر کی جھونپڑی مذبح خانے کے پاس تھی ۔ جانور ذبح ہوتے وقت درد بھری آواز نکالتے تو آپ پریشان ہوکر کلی سے باہر آجاتے ۔ آخر آپ نے فرمایا :
کبیرا تیری جھونپڑی گل کٹیوں کے پاس
جو کرے سو بھرے توں کیوں رہے اداس
یہ دنیا کی زندگی بہت ہی مختصر ہے۔ سب کچھ یہیں چھوڑ کر چلے جانا ہے۔ کسی نے کیا خوب لکھا ہے:
اجڑ گئے وہ باغ جس کے لاکھوں مالی تھے
سکندر جب گیا دنیا سے دونوں ہاتھ خالی تھے
لکھیا چن چراغ نے سرفراز خان نوں:
کجھ کھا لیئے کجھ پی لیئے کچھ دے دیئے رحمان نوں
اتنا غرور نہیں چاہیے اس قبر دے مہمان نوں
چن چراغ میاں اس پینگ نے ٹٹ جاونا
ایہ جیہڑی چڑھی ہے اسمان نوں
اس زندگی کو کیوں نہ اللہ کی یاد میں...

پاکستانی معاشرے میں طلاق کا بڑھتا ہوا رجحان اور اس کے اسباب

Islam wants from its believers to make a peaceful society. The first base of each society is husband-wife relation. Islam has given much emphasis upon this relationship to make it smooth, peaceful, joyful and interactive. But considering human as multidimensional, Islam has allowed husband and wife to get themselves separate from each other, it they cannot survive this relationship smoothly at any level. Though, ‘divorce’ is allowed in Islam but at last solution. Pakistan, as being a Muslim society is facing increase rate in divorce nowadays. My research work is covering different reasons and aspects behind this high ratio of divorce in Pakistan. This research will be helpful to find out any solution to decrease the divorce ration in Pakistani society.

Fabrication of Al-Cnts Mmc by Induction Melting With Improved Dispersion and Wetting of Cnts in Al Matrix Using a Multifunctional Flux

Aluminum/carbon nanotubes (Al-CNTs) composite is an encouraging candidate material for aerospace applications due to its expected high strength-to-weight ratio. Carbon nanotubes (CNTs) offer remarkable reinforcements owing to their high specific strength and specific modulus. However, uniform dispersion and wetting of the CNTs is extremely difficult in molten aluminum, due to large difference in surface tension forces of the two components. In present work, the dispersion issue was improved using induction melting technique, where innate stirring action of induction melting dispersed the nanotubes in molten aluminum. The wetting was improved using a multifunction flux (titanate of potassium), which, when was incorporated in molten mixture of aluminum and CNTs having, instigated in-situ reactions to form titanium carbides on the surface of the nanotubes causing increased wetting of CNTs by molten aluminum. The composites were characterized using scanning electron microscopy, x-ray diffraction, transmission electron microscopy and mechanical testing. Refinement in crystallite size was achieved down to ~150 nm and a corresponding increase in lattice strain up to ~3.46x10-3 was observed in the composites. A simultaneous increase in v i yield strength ~208 %, tensile strength ~218 %, and hardness ~100 % was observed. However, the decrease in the ductility of the composite associated with the strengthening of the matrix was <25 %. Additionally, stress relaxation behavior of the annealed composite was improved by ~30 % compared with pure aluminum. Consequently, the stress relaxation rate of the composite was decreased even beyond the yield strength of the annealed pure aluminum. Therefore, induction melting and usage of the flux for improvement in the dispersion and wetting of the nanotubes, respectively, appeared to be a potential method to fabricate Al-CNTs composites.