kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Evaluation of AI generated ligands for bioprocess application
KTH, School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH), Protein Science.
2024 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Utvärdering av AI-genererade ligander för bioprocesstillämpning (Swedish)
Abstract [sv]

Integrationen av artificiell intelligens (AI) i bioprocessapplikationer har framträtt som en transformerande metod inom bioteknik- och läkemedelsindustrin, och lovar betydande framsteg i effektivitet, verkan och hållbarhet. Genom att utnyttja avancerade algoritmer, såsom djupinlärning och förstärkningsinlärning, kan AI-system förutsäga och generera nya ligandstrukturer med hög affinitet för sina mål, såsom monoklonala antikroppar (mAbs) i detta projekt.

Detta projekt syftar till att utvärdera potentialen hos AI-genererade affinitetsproteiner och validera de datorsimulerade förutsägelserna genom att undersöka bindningseffektiviteten och stabiliteten hos dessa ligander under verkliga förhållanden. Flera våtlabbstekniker användes för att uttrycka och rena de AI-designade proteinerna. Affinitetskromatografi var en teknik som användes för rening, följt av ytplasmonresonans (Biacore) för att studera interaktionen mellan de genererade affinitetsproteinerna och mAbs.

Analyseresultat från SDS-PAGE och masspektrometri visade att de flesta proteiner kunde renas med hjälp av affinitetskromatografi. Emellertid visade karaktärisering med Biacore att de flesta proteiner inte interagerade med mAbs, förutom ett designat protein. Cirkulär dikroism (CD) spektrometri som användes för att visualisera sekundärstrukturen i proteiner visade att de flesta proteiner var veckade och bibehöll alfahelixar och betaflak jämfört med det vilda typen proteinet.

Sammanfattningsvis ger denna forskning värdefulla insikter i utmaningarna vid utvärdering och karaktärisering av AI-genererade proteiner. Ytterligare forskningsinsatser bör fokusera på att förfina experimentella förhållanden och visualisera sekundärstrukturerna hos de genererade proteinerna för en djupare förståelse av deras stabilitetsproblem.

Abstract [en]

The integration of artificial intelligence (AI) into bioprocess applications has emerged as a transformative approach in the biotechnology and pharmaceutical industries, promising significant advancements in efficiency, efficacy, and sustainability. By leveraging advanced algorithms, such as deep learning and reinforcement learning, AI systems can predict and generate novel ligand structures with high affinity for their targets, such as monoclonal antibodies(mAbs) in this project.

This project aims to evaluate the potential of AI-generated affinity protein ligands and validate the computational predictions by examining the binding efficiency and stability of these ligands in real-world conditions. Several wet lab techniques were employed to express and purify the AI-designed proteins. Affinity chromatography was one technique used for purification, followed by surface plasmon resonance (Biacore) to study the interaction between the generated affinity proteins and mAbs.

Analysis results from SDS-PAGE and mass spectrometry showed that most proteins could be purified using affinity chromatography. However, characterization using Biacore revealed that most proteins did not interact with mAbs, except for one designed protein. Circular dichroism (CD) spectrometry used to visualize the secondary structure in proteins showed that most proteins were folded and retained alpha helices and beta sheets when compared to the wild type protein.

In conclusion, this research provides valuable insights into the challenges in evaluating and characterizing AI-generated proteins. Further research efforts should focus on refining experimental conditions and visualizing the secondary structures of the generated proteins for a more in-depth understanding of their stability issues.

Place, publisher, year, edition, pages
2024.
Series
TRITA-CBH-GRU ; 2024:242
Keywords [en]
affinity proteins, AI tools, affinity chromatography, binding interaction, monoclonal antibodies
Keywords [sv]
affinitets proteiner, AI-verktyg, affinitetskromatografi, bindningsinteraktion, monoklonala antikroppar
National Category
Medical Biotechnology (with a focus on Cell Biology (including Stem Cell Biology), Molecular Biology, Microbiology, Biochemistry or Biopharmacy)
Identifiers
URN: urn:nbn:se:kth:diva-348476OAI: oai:DiVA.org:kth-348476DiVA, id: diva2:1876278
External cooperation
Cytiva
Subject / course
Biotechnology
Educational program
Master of Science - Medical Biotechnology
Supervisors
Examiners
Available from: 2024-06-24 Created: 2024-06-24

Open Access in DiVA

fulltext(10765 kB)838 downloads
File information
File name FULLTEXT01.pdfFile size 10765 kBChecksum SHA-512
374a75aa9df0752780b199549ece5c3d14872987235e2201cf0822595937d1bfb6b10807a5cbf935ca5dca9c50bf4fde1c13981dae8b4160df532e4589fcc398
Type fulltextMimetype application/pdf

By organisation
Protein Science
Medical Biotechnology (with a focus on Cell Biology (including Stem Cell Biology), Molecular Biology, Microbiology, Biochemistry or Biopharmacy)

Search outside of DiVA

GoogleGoogle Scholar
Total: 838 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 599 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf