Daniella Pretorius
@daniellapret
Followers
198
Following
600
Media
9
Statuses
100
🇿🇦 | PhD Candidate at @imperialcollege in the @jwm_imperial lab | 🔬 Protein Design | 🧬 Biochemistry | 🤖 Explainable AI | 🌐
Joined February 2021
🚨 Thrilled to share one of my main PhD projects!.We built an in silico evolution platform that couples a solenoid discriminator network with AlphaFold2 as an oracle, using a genetic algorithm for sequence update. 🧬✨ (1/9)
2
27
137
RT @JeffDean: Colab Pro is now available for free for verified US students and faculty for one year.
0
81
0
RT @BiologyAIDaily: Designing Novel Solenoid Proteins with In Silico Evolution. 1. This study presents a groundbreaking in silico evolution….
0
9
0
Massive thanks to my co-authors and everyone who supported this project! 🙌 @jwm_imperial, @G__Nikov @sokrypton and everyone not on X (9/9).
0
0
6
Overall, we experimentally characterised 41 solenoid designs. I really love this class of protein ❤️, so super happy to put this into the world (below, me and 25 overnight cultures from the first round!) (8/9)
1
0
6
We refined our design filters and added capping regions in a second round of β-solenoid designs. This led to constructs with biophysical properties matching their design models. (7/9)
2
0
5
In our first experimental round, the all α-helical α-solenoid designs expressed and folded. We even solved a crystal structure that closely matched our design model! But all β-solenoid designs failed. (6/9)
1
0
3
All designed β-solenoids were right-handed, despite both left- and right-handed versions existing in nature. We observed this often in the AlphaFoldDB, hinting at a possible model bias for certain repeat-rich sequences. (5/9)
1
0
6
Starting from random sequences, we generated α-, β-, and αβ-solenoids that cover both natural and novel regions of structure space. To our knowledge, this is the first fold-specific, hallucination based design method that doesn’t rely on structural templates. (4/9)
1
0
6
Why solenoids? These tandem repeat proteins have very diverse roles in nature - antifreeze proteins and DNA binders to name a few (3/9)
1
0
4
RT @ml4proteins: Next Tues (4/29) at **4:30PM** ET, we will have.@ginaelnesr @HWaymentSteele present "Learning millisecond protein dynamics….
biorxiv.org
Many proteins’ biological functions rely on interconversions between multiple conformations occurring at micro-to millisecond (µs-ms) timescales. A lack of standardized, large-scale experimental data...
0
12
0
RT @ChoYehlin: Excited to share our preprint “BoltzDesign1: Inverting All-Atom Structure Prediction Model for Generalized Biomolecular Bind….
0
116
0
RT @_JosephWatson: I’m excited to share our significantly-updated preprint on de novo antibody design, where we now demonstrate the structu….
0
68
0
RT @brianltrippe: 🔥 Benchmark Alert! MotifBench sets a new standard for evaluating protein design methods for motif scaffolding. Why does t….
0
41
0
RT @KevinKaichuang: A protein language model trained to predict subcellular localization for human proteins can generate de novo sequences….
0
37
0
RT @ChoYehlin: Thrilled to share my recent work with @sokrypton, @grocklin, @KotaroTsuboyama, and @JustasDauparas: .
biorxiv.org
Generative protein modeling provides advanced tools for designing diverse protein sequences and structures. However, accurately modeling the conformational landscape and designing sequences—ensuring...
0
24
0
RT @NerdyChristie: Small but mighty: The differences in these two butterflies' wings are driven by a microRNA, a discovery that solves a lo….
0
11
0
RT @chrisfrank662: After showing that AF2 can be used to design very large proteins by performing #RSO, I am happy to share another fun pro….
biorxiv.org
Recent advances in computational protein design have enabled the creation of novel proteins for a variety of purposes. The capability for producing custom-shape high-quality backbones for very large...
0
47
0
RT @liambai21: Ever wondered how a protein language model sees your favorite protein?. Checkout out our SAE visualizer where you can now se….
0
46
0