File size: 1,520 Bytes
0b479b2
bb0614c
 
 
7b41737
bb0614c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
# Trastuzumab-scFv CDR3 Optimization Variants Iter0 x HER-2 (YM_0989)

## Overview

YM_0989 are Trastuzumab designs against HER2. We explored several model hypothesis: (i) Does pre-training aid predicitivity and (ii) does the featurization of the input sequences matter. To test pretraining, we refer to `mata_descriptions` with the term **warm** to include pretraining, and **cold** to start from a randomly initialized seed. For featurization, we explored **label-encoded** sequences with a one-hot-encoder of amino acid identities, versus an **ESM-featurized** embedding to represent each sequence in the PPI.

## Experimental details

We studied the efficacy of generating binders with different model hyperparameters. This dataset includes 20828 unique VHHs and 1 unique sequences. All designs are limited to the CDRs of the proteins of interest

A more extensive methods section can be found in our publication [here](https://pmc.ncbi.nlm.nih.gov/articles/PMC12296056/).

## Misc dataset details

We define the following binders:

### A-library (scFvs)

There are several terms you can filter by:

- `TrastuzumabCDR_WT_<i>`: These are WT replicates.
- `TrastuzumabCDR_label_encoded_cold`: Label encoded sequences with no pretraining
- `TrastuzumabCDR_label_encoded_warm`: Label encoded sequences with pretraining
- `TrastuzumabCDR_esm_cold`: ESM featurized sequences with no pretraining
- `TrastuzumabCDR_esm_warm`: ESM featurized sequences with pretraining

### Alpha-library

There is only 1 target of interest.