iiinova — where your digital self keeps evolving

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

Working with @aviezerl.bsky.social has been a blast, the wonderful Roni Stok & Saifeng Cheng (with Yonatan Stelzer’s guidance) conducted the experiments and data collection, with all efforts orchestrated and led by Amos Tanay. 14/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

IQ lays a foundation for modeling other epigenomic features. We're using it to study Polycomb domains, DNA methylation and insulation. We optimistically believe all epigenomic aspects can be combined to model differentiation programs from sequence alone. 13/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

We are also curious to learn new biology when black box models outperform IQ. In CRE sequences alone, LLMs have not yet revealed grammars that transcend IQ’s simple TF-DNA interactions. Longer-range chromosomal interactions among CREs, we believe, may be a different story. 12/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

We are excited about the possibility of using IQ to develop better DL for epigenomes. We do gain from using ensembles of IQ and DL models – so there is hope! LLMs consider larger contexts than IQ, while IQ is super economical in parameters and may generalize better because of that. 11/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

Insight 3: Local interactions between TFs. IQ predicts TF interactions that predict CRE accessibility across differentiation trajectories. For example: Mesp-Eomes motif co-occurences may be important for germ-layer specification. 10/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

Insight 2: TFs care about sub-optimal binding sites. IQ integrates strong and weak TF-DNA interactions. This reveals that TFs 'read' sequences in different ways – some care only for the best targets and others integrate many weak ones. 9/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

Here are 3 biological insights derived from IQ (before we reflect on what’s in it for the non-biologist crowd). Insight 1: Sequence defines regulatory intensity over a quantitative spectrum – not a binary yes/no classifier. IQ can predict this spectrum! 8/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

For example, when modelling regulation of Epiblast to Mesoderm differentiation in mouse embryos, we can explain changes in accessibility with only 13 TF motif models! 7/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

IQ inference starts from detailed model and progressively simplifies it to a small set of physical models. Models with smaller number of components are easier to interpret. They also generalize better. 6/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

IQ regresses AP from sequence using biophysically inspired TF binding models including: *Spatial integration across a range of TF-DNA affinities *Latent TF concentrations with non-linear dose-response *Synergistic/antagonistic pairwise TF interactions 5/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

IQ uses a normalization trick based on coverage of constitutive ATAC peaks to derive APs robustly. Because AP is defined on an absolute scale – comparing AP among conditions is immediate and robust. 4/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

But IQ is more than a predictive tool. A first important difference between IQ and other approaches is the transformation of ATAC-seq coverage to access probabilities (AP) - the instantaneous chances to find a CRE in an open state among cells from a given type. 3/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

As a mere predictive tool, IQ performs on par with state-of-the-art DL models such as Borzoi and DeepTopic. Evaluation on human blood and mouse embryo datasets shows that current best performance is derived using an ensemble of IQ and Borzoi. 2/

100

akhiad.bsky.social @akhiad.bsky.social · 17/10/2025

Do we need LLMs to predict epigenomes from DNA—or is biophysics enough? 🧬 IceQream (IQ) is a biophysics-based framework that predicts epigenomes with SOTA-level accuracy—and is fully explainable. @NatureComms: www.nature.com/articles/s41... Thread 🧵👇 1/

1184