Lightnews — Scholar-powered news

Reposted by Kuan-Hao Chao

bioRxiv Genomics @biorxiv-genomic.bsky.social · 19d

Predicting dynamic expression patterns in budding yeast with a fungal DNA language model https://www.biorxiv.org/content/10.1101/2025.09.19.677475v1

1 1

Reposted by Kuan-Hao Chao

JHU Computer Science @jhucompsci.bsky.social · Sep 4

Congratulations, @kuanhaochao.bsky.social!

Congratulations! Kuan-Hao Chao successfully defended his dissertation “Decoding the Language of Genomes: Bridging Sequences and Function Through Deep Learning” under the guidance of advisors Steven Salzberg and Mihaela Pertea. Kuan-Hao plans to join Illumina’s artificial intelligence lab as a senior deep learning scientist. We in the department are extremely proud of our students who have successfully completed their PhD. Congratulations on this achievement and best wishes as you begin an exciting new phase of life!

1 4

Kuan-Hao Chao @kuanhaochao.bsky.social · Aug 26

Heart full—I defended my PhD in @jhucompsci.bsky.social @jhu.edu(Aug 25). Thanks to my advisors @stevensalzberg.bsky.social and Mihaela Pertea; my committee— @benlangmead.bsky.social, David Kelley & Anqi Liu; and all labmates, collaborators, and mentors. You made me who I am today. Deeply grateful!

4 15

Kuan-Hao Chao @kuanhaochao.bsky.social · Aug 26

Thank you so much Ben. I’m deeply grateful for your mentorship; you’ve shaped how I think and do science. I hope we get to collaborate again soon!

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Aug 25

It’s happening today! Thrilled to share my PhD journey with friends and colleagues—thank you all for being part of it!

1 3

Kuan-Hao Chao @kuanhaochao.bsky.social · Aug 21

Friends I’m defending my PhD!

🗓 Mon, Aug 25 • 2:30 PM–4:30 PM ET
📍 JHU Homewood — Clark Hall 110
💻 Zoom: jhubluejays.zoom.us/j/9385611117...

If you’re into genomics or deep learning (or just curious how AI reads DNA), I’d love to have you there—friends welcome!
Details: khchao.com/thesis-defen...

Join our Cloud HD Video Meeting

Zoom is the leader in modern enterprise cloud communications.

jhubluejays.zoom.us

1 4

Kuan-Hao Chao @kuanhaochao.bsky.social · Aug 5

Thrilled to join Illumina AI Lab as Senior Deep Learning Engineer! Huge thanks to my mentors & friends at @jhu.edu and Calico, and a special shout-out to Kyle, Kishore, and Tracie for the warm welcome. Excited to build AI-driven genomics solutions together! #AI #Genomics

I’m thrilled to announce that I’ve joined the Illumina Artificial Intelligence Lab as a Senior Deep Learning Scientist & Engineer! | Kuan-Hao Chao

I’m thrilled to announce that I’ve joined the Illumina Artificial Intelligence Lab as a Senior Deep Learning Scientist & Engineer! 🎉 It’s been an absolute privilege to collaborate with the talented ...

www.linkedin.com

2 1 4

Kuan-Hao Chao @kuanhaochao.bsky.social · Jul 23

Dive into the full thread here 👇

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Friends, I’m excited to release OpenSpliceAI, an open‐source, efficient, and modular framework for splice site prediction. It reimplements and extends SpliceAI (Jaganathan et al., 2019) using the modern PyTorch framework. 1/

Mihaela Pertea, @stevensalzberg.bsky.social, Anqi Liu, Alan Mao

bioRxiv Genomics @biorxiv-genomic.bsky.social · Mar 24

OpenSpliceAI: An efficient, modular implementation of SpliceAI enabling easy retraining on non-human species https://www.biorxiv.org/content/10.1101/2025.03.20.644351v1

Kuan-Hao Chao @kuanhaochao.bsky.social · Jul 23

🚀 Thrilled to share our OpenSpliceAI paper is out in @elife.bsky.social
Open‑source, PyTorch extension of SpliceAI—fast, modular splice‑site prediction, transfer learning, and variant‑effect scoring.

🔗 Paper doi.org/10.7554/eLif...
🔗 Code github.com/Kuanhao-Chao...
🔗 Docs ccb.jhu.edu/openspliceai/

1 2 10

Kuan-Hao Chao @kuanhaochao.bsky.social · Jun 17

Excited to introduce LiftOn – an open-source tool for accurate, scalable liftover of genome annotations (GFF) across assemblies. 🚀

👉 Code & community: github.com/Kuanhao-Chao...

It’s been incredibly rewarding building this for the genomics community. Can’t wait for your feedback and contributions!

8 22

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 25

@jhucompsci.bsky.social @hopkinsengineer.bsky.social

1

Reposted by Kuan-Hao Chao

Steven Salzberg @stevensalzberg.bsky.social · Mar 24

@kuanhaochao.bsky.social and Alan Mao built an open-source version of SpliceAI, a deep learning model for predicting human splice sites, and it turned into quite a project! Many new features including training modules, check out the whole thread below

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Friends, I’m excited to release OpenSpliceAI, an open‐source, efficient, and modular framework for splice site prediction. It reimplements and extends SpliceAI (Jaganathan et al., 2019) using the modern PyTorch framework. 1/

Mihaela Pertea, @stevensalzberg.bsky.social, Anqi Liu, Alan Mao

bioRxiv Genomics @biorxiv-genomic.bsky.social · Mar 24

OpenSpliceAI: An efficient, modular implementation of SpliceAI enabling easy retraining on non-human species https://www.biorxiv.org/content/10.1101/2025.03.20.644351v1

1 3

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

A huge thank you to my advisors, Mihaela Pertea and @stevensalzberg.bsky.social for their invaluable mentorship and support. Special shoutout to collaborators and friends Anqi Liu and Alan Mao at Hopkins — this project wouldn’t be possible without you! 11/

1 1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

OpenSpliceAI offers researchers a comprehensive suite of tools for studying transcript splicing—from creating training datasets and training models to predicting splice sites and assessing the impact of genetic variants.
🔗 Explore our documentation here: ccb.jhu.edu/openspliceai/ 10/

1 1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

We showed general patterns of donor and acceptor sites across hundreds of splice site motifs learned by OpenSpliceAI. Moreover, it confidently predicts cryptic splicing events—such as acceptor gain in MYBPC3 and novel exon gain in OPA1. 9/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

In silico mutagenesis (ISM) confirms that OpenSpliceAI focuses on the same key regions and patterns for splice site prediction at U2SURP and DST as SpliceAI, and effectively capturing a splicing enhancer. We also demonstrate its ability to capture the full gene span of CFTR. 8/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

We enhanced model reliability by applying temperature scaling for calibration. Using expected calibration error and reliability diagrams, we observed a smoother probability distribution that aligns more closely with the empirical distribution, making it better! 7/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

OpenSpliceAI also supports transfer learning. Our experiments show that models pre-trained on human and applied to species of interest achieve near-optimal performance in just one epoch, drastically reducing compute time while enhancing predictions for species with smaller genomes 6/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Comparing species-specific training versus running SpliceAI directly, our results confirm that one-size-fits-all generalization isn’t enough. OpenSpliceAI’s ability to retrain on specific species sets it apart from SpliceAI. 5/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Our benchmarks show that OpenSpliceAI outperforms SpliceAI in elapsed time, memory usage, and GPU peak memory across various gene lengths. Thanks to dynamic PyTorch graphs, batch prediction, and optimized engineering, full-chromosome predictions are now a reality! 4/

1 1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Built with six modular components, OpenSpliceAI lets you:

• Create species-specific datasets
• Train custom models
• Calibrate predictions
• Apply transfer learning from human models
• Predict on genes / entire chromosomes
• Assess variant impacts on cryptic splicing
3/

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

We’re introducing OpenSpliceAI along with a suite of pre-trained models for Human-MANE, mouse, zebrafish, honeybee, and Arabidopsis. OpenSpliceAI provides a user‐friendly toolkit for studying transcript splicing in any species of interest! 2/

🔗 GitHub: github.com/Kuanhao-Chao...

GitHub - Kuanhao-Chao/OpenSpliceAI: 🤖 Open‑source deep-learning-based splice‑site predictor that decodes splicing patterns across species

🤖 Open‑source deep-learning-based splice‑site predictor that decodes splicing patterns across species - Kuanhao-Chao/OpenSpliceAI

github.com

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Friends, I’m excited to release OpenSpliceAI, an open‐source, efficient, and modular framework for splice site prediction. It reimplements and extends SpliceAI (Jaganathan et al., 2019) using the modern PyTorch framework. 1/

Mihaela Pertea, @stevensalzberg.bsky.social, Anqi Liu, Alan Mao

bioRxiv Genomics @biorxiv-genomic.bsky.social · Mar 24

OpenSpliceAI: An efficient, modular implementation of SpliceAI enabling easy retraining on non-human species https://www.biorxiv.org/content/10.1101/2025.03.20.644351v1

1 7 24

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 24

Friends, I’m excited to release OpenSpliceAI, an open‐source, efficient, and modular framework for splice site prediction. It reimplements and extends SpliceAI (Jaganathan et al., 2019) using the modern PyTorch framework. 1/
Mihaela Pertea, @stevensalzberg.bsky.social, Anqi Liu, Alan Mao

1

Kuan-Hao Chao @kuanhaochao.bsky.social · Mar 3

I want to again thank my advisors @stevensalzberg.bsky.social and Mihaela Pertea for their incredible support and mentorship. A big shoutout to all my collaborators and friends at Hopkins as well! 8/8

1