TauAge

TauAge model and its significance in understanding regional tau pathology in aging, Alzheimer's disease (AD), and Primary Age-Related Tauopathy (PART):

---

🧬 What is TauAge?

TauAge is a DNA methylation (DNAm)-based epigenetic "clock" created by the authors to estimate the severity of phosphorylated tau (p-tau) pathology in specific brain regions.

Unlike traditional epigenetic clocks (e.g., Horvath clock) that estimate biological age, TauAge is trained to predict the pathological burden of p-tau, adjusted for chronological age. It provides a molecular readout of tau accumulation independent of aging.

---

⚙️ How TauAge Was Built

1. Input Data:

   • DNA methylation profiles (DNAm) from the frontal cortex (Illumina arrays).
   • Corresponding histological quantification of p-tau in either the hippocampus or frontal cortex (depending on model).
   • Two cohorts: PWG (PART-only) and ROSMAP (AD and PART).

2. Modeling Approach:

   • The authors used elastic net regression (a regularized machine learning model) to predict age-adjusted residuals of p-tau.
   • This means they removed the effect of chronological age to focus on DNAm markers that predict tau pathology beyond what you'd expect for someone’s age.

3. Output:

   • A TauAge score per sample, reflecting how much tau burden is present in a region (hippocampus or frontal cortex), based purely on DNAm.
   • Each model uses hundreds of CpG sites as features.

---

🧠 Why Separate Models for Hippocampus and Frontal Cortex?

• Tau pathology originates in the hippocampus (early) and spreads to the frontal cortex (later) in AD.
• PART, by contrast, typically remains confined to the hippocampus and does not spread to cortex.

Thus, the biology driving p-tau accumulation is likely region-specific. The authors found:

• Only 8 CpGs overlapped between the hippocampal and frontal TauAge models.
• The models were not interchangeable — e.g., using hippocampal CpGs to predict frontal tau gave poor results.

➡️ Conclusion: p-tau pathology is regulated by distinct molecular programs in different brain regions.

---

🧠 What Does TauAge Reveal About AD vs. PART?

1. Hippocampal TauAge (PART & AD)

• Biological associations:

  • Genes linked to TauAge CpGs are enriched in:
    • Synaptic transmission
    • Ion transport
    • Mitochondrial function
  • These functions are shared in both PART and AD.

• Interpretation:

  • Tau accumulation in the hippocampus is driven by synaptic dysfunction, regardless of whether it's PART or AD.
  • Suggests that PART and early AD may share early tau-related vulnerability via similar epigenetic pathways.

---

2. Frontal Cortex TauAge (Only in AD)

• Biological associations:

  • CpG-correlated genes are enriched in:
    • Inflammatory signaling
    • Cytokine production
    • Immune cell recruitment (e.g., T-cells, macrophages)

• Interpretation:

  • Tau pathology in the frontal cortex (seen in later AD) is associated with inflammation, which is absent in PART.
  • This provides molecular evidence that neuroinflammation is a distinguishing feature of AD, not PART.

---

🧪 Validation and Predictive Power

• TauAge outperformed traditional clocks like the Horvath pan-tissue clock or the brain-specific Cortical clock in predicting tau burden.
• It accurately tracked regional pathology and even distinguished:
  • Mild Cognitive Impairment (MCI) vs. Dementia.
  • Cognitive resilience in Predicted-PART vs. Predicted-AD groups.

---

💡 Why TauAge Matters

1. Region-specific insight: First model to quantify tau pathology in specific brain areas using methylation.
2. Separates aging from disease: Measures pathology beyond what’s expected from age alone.
3. Molecular divergence of PART vs. AD: Synaptic dysfunction is shared; inflammation is unique to AD.
4. Biomarker potential: Could be developed into a diagnostic or risk-stratifying tool, especially if adapted to blood or CSF methylation data.
5. Framework for further modeling: Could be extended with scRNA-seq, single-cell methylomics, or used in AI-based diagnostic pipelines.

---

Would you like a visual summary or help implementing a TauAge-like model on your own data?

attack strategies targeting reward functions in reinforcement learning (RL) AI models,

 To systematically explore attack strategies targeting reward functions in reinforcement learning (RL) AI models, the following research strategy integrates theoretical analysis, empirical validation, and mitigation development:

1. Problem Definition & Vulnerability Analysis

• Objective: Identify how reward function design influences exploitability.

• Approach:

  • Classify reward functions by type (extrinsic/intrinsic/shaped1), structure (sparse/dense17), and optimization goals (cost minimization vs. reward maximization1).

  • Analyze attack surfaces:

    • Reward hacking: Exploiting unintended loopholes (e.g., infinite reward loops4).

    • Reward poisoning: Manipulating training data or reward signals to induce malicious policies56.

  • Prioritize vulnerabilities in adaptive vs. non-adaptive attacks, noting that adaptive methods can achieve goals faster6.

2. Attack Methodology Development

• Objective: Design and benchmark attack vectors.

• Approach:

  • Poisoning Attacks:

    • Training-time: Inject adversarial perturbations into reward signals (e.g., δₜ in rt + δₜ6).

    • Preference-based: Manipulate human feedback datasets (e.g., label-flipping in RLHF5).

  • Exploratory Attacks:

    • Test reward-shaping vulnerabilities (e.g., sparse reward exploitation7).

    • Develop adaptive attacks leveraging agent policy updates during training6.

  • Metrics: Measure attack success via policy divergence, time-to-compromise, and stealthiness (e.g., detectability of manipulated rewards5).

3. Empirical Validation

• Objective: Evaluate attacks across RL environments.

• Testbeds:

  • Simulations: Gridworlds, MuJoCo, Atari games.

  • Real-world tasks: Autonomous navigation, recommendation systems.

• Variables:

  • Compare attack efficacy under different reward structures (e.g., dense vs. sparse17).

  • Test robustness of model-based vs. model-free RL algorithms.

4. Impact Analysis

• Objective: Quantify consequences of compromised reward functions.

• Key Questions:

  • How do poisoned rewards degrade policy performance or safety?

  • Can attacks induce catastrophic forgetting or goal hijacking?

  • What is the minimum perturbation (δₜ) required for successful attacks6?

5. Defense Mechanisms

• Objective: Propose countermeasures.

• Strategies:

  • Robust Reward Design:

    • Uncertainty-aware reward inference (e.g., MURAL7).

    • Constrained optimization to limit reward manipulation3.

  • Detection Systems:

    • Anomaly detection in reward distributions.

    • Adversarial training with perturbed rewards6.

  • Formal Guarantees: Certify safety thresholds for reward perturbations6.

6. Ethical & Practical Considerations

• Objective: Ensure responsible research practices.

• Guidelines:

  • Restrict testing to controlled environments.

  • Collaborate with AI safety communities to preempt real-world misuse.

  • Disclose vulnerabilities to affected frameworks (e.g., OpenAI, DeepMind).

Implementation Timeline

Phase	Duration	Deliverables
1–2	6 months	Taxonomy of vulnerabilities; attack prototypes
3–4	9 months	Benchmark results; impact analysis framework
5–6	12 months	Defense toolkit; ethical guidelines

This strategy balances offensive exploration (to identify risks) and defensive innovation (to mitigate harms), advancing both RL security and robustness.

Citations:

1. https://www.linkedin.com/pulse/rewards-reinforcement-learning-caleb-m-bowyer
2. https://spinningup.openai.com/en/latest/spinningup/rl_intro.html
3. https://proceedings.neurips.cc/paper_files/paper/2021/file/a7f0d2b95c60161b3f3c82f764b1d1c9-Paper.pdf
4. https://rodtrent.substack.com/p/must-learn-ai-security-part-12-reward
5. https://aclanthology.org/2024.acl-long.140.pdf
6. https://par.nsf.gov/servlets/purl/10183713
7. http://bair.berkeley.edu/blog/2021/10/22/mural/
8. https://learnprompting.org/blog/openai-solution-reward-hacking
9. https://jair.org/index.php/jair/article/view/12440
10. https://openreview.net/forum?id=25G63lDHV2
11. https://isaacperper.com/images/6881_intrinsic/6_881_Final_Report.pdf
12. https://www.alignmentforum.org/posts/Ge55vxEmKXunFFwoe/reward-hacking-behavior-can-generalize-across-tasks
13. https://par.nsf.gov/servlets/purl/10183709
14. https://alignment.anthropic.com/2025/reward-hacking-ooc/
15. https://lilianweng.github.io/posts/2024-11-28-reward-hacking/
16. https://www.reddit.com/r/reinforcementlearning/comments/1ae9t90/how_does_reward_work_while_training_a/
17. https://arxiv.org/abs/2205.15400
18. https://en.wikipedia.org/wiki/Reinforcement_learning
19. https://www.reddit.com/r/reinforcementlearning/comments/vicory/does_the_value_of_the_reward_matter/
20. https://ojs.aaai.org/index.php/AAAI/article/view/26240/26012
21. https://arxiv.org/pdf/2102.08492.pdf
22. https://www.semanticscholar.org/paper/Reward-Machines:-Exploiting-Reward-Function-in-Icarte-Klassen/6778d6a0f959cdcc42718ee9fc279fd1f00f3d88
23. https://www.sciencedirect.com/science/article/abs/pii/S0925231223007014
24. https://arxiv.org/abs/2211.09019
25. https://arxiv.org/html/2402.09695v1
26. https://stackoverflow.com/questions/47133913/what-is-importance-of-reward-policy-in-reinforcement-learninig
27. https://deepblue.lib.umich.edu/bitstream/handle/2027.42/136931/guoxiao_1.pdf
28. https://stats.stackexchange.com/questions/189067/how-to-make-a-reward-function-in-reinforcement-learning
29. https://ai.stackexchange.com/questions/22851/what-are-some-best-practices-when-trying-to-design-a-reward-function
30. https://www.reddit.com/r/reinforcementlearning/comments/12jey74/exploiting_the_model_in_reinforcement_learning/
31. https://github.com/RodrigoToroIcarte/reward_machines
32. https://www.jair.org/index.php/jair/article/download/12440/26759/29354

---

Answer from Perplexity: pplx.ai/share

Targeting Tau in Alzheimer's Disease

pipp1 2525493

 pipp1 2525493

https://www.nsf.gov/awardsearch/showAward?AWD_ID=2525493&HistoricalAwards=false

The Religious Orders Study (ROS) and the Rush Memory and Aging Project (MAP), collectively known as ROSMAP,

 The Religious Orders Study (ROS) and the Rush Memory and Aging Project (MAP), collectively known as ROSMAP, are two longitudinal clinical-pathologic cohort studies aimed at understanding aging and Alzheimer's disease (AD). Both studies are conducted by the Rush Alzheimer's Disease Center at Rush University Medical Center in Chicago, Illinois.

Religious Orders Study (ROS): Initiated in 1994, ROS enrolls Catholic nuns, priests, and brothers aged 65 and older from across the United States. Participants without known dementia agree to annual medical and psychological evaluations and consent to brain donation after death. To date, approximately 1,200 individuals have enrolled, with around 580 currently alive. Yearly blood draws are also conducted, resulting in the storage of serum, plasma, and cells. citeturn0search0

Rush Memory and Aging Project (MAP): Launched in 1997, MAP recruits older individuals from about 40 continuous care retirement communities and senior subsidized housing facilities around the Chicago metropolitan area. Similar to ROS, participants without known dementia undergo detailed annual clinical evaluations and agree to donate their brain, spinal cord, and muscle tissues after death. Over 1,600 older adults have enrolled, with approximately 1,000 currently alive. Yearly blood draws are also part of the study protocol. citeturn0search0

Study Objectives and Methodology: Both ROS and MAP aim to identify risk factors for cognitive decline and incident AD dementia, among other health outcomes. Participants undergo annual detailed clinical evaluations, including cognitive performance testing, self-reported assessments, and medication inspections to document medical conditions. A decision tree designed to mimic expert clinical judgment is implemented by computer to inform several clinical diagnoses, including dementia and AD. citeturn0search0

Contributions to Alzheimer's Disease Research: ROSMAP has significantly advanced the understanding of the mechanisms underlying neurodegenerative diseases, particularly AD. The studies have provided insights into the relation of neurobiology to clinical AD, neurobiologic pathways linking risk factors to clinical AD, non-cognitive AD phenotypes including motor function and decision making, and the development of novel drug discovery platforms. citeturn0search1

In summary, ROSMAP serves as a valuable resource for the aging and dementia research community, offering a wealth of data and findings that can be leveraged for future studies.

AD Knowledge portal

 

https://adknowledgeportal.synapse.org/Explore/Studies/DetailsPage/StudyDetails?Study=syn18485175

study

The Single-cell transcriptomic analysis of Alzheimers disease (snRNAseqPFC_BA10) Study

Harvard University, Massachusetts Institute of Technology, Rush University

This study provides single-cell transcriptomic data from the prefrontal cortex from individuals sleceted from the Religious Orders Study and Memory and Aging Project (ROSMAP).

Data Types	Gene Expression
Study Focus	Alzheimer Disease
Specimen Type	Brain
Species	Human
Program	CDCP
Grant	RF1AG054321

from bulk sequence to scRNA

 bulk RNAseq is cheap and easy to geneerate. scRNA is expensive but more inforamtive. 

https://github.com/AI-in-Biomedical-Science/scaden

2025 0307 Friday AAAI review

 Zoom, start recording

AAAI review (underline) 

Andrew Ng (talk), other talks

https://underline.io/events/473/sessions/19421/lecture/115522-ai-agents-and-applications

TH05: Multi-modal Foundation Model for Scientific Discovery: With Applications in Chemistry, Material, and Biology

AI for science: 

https://zoom.us/rec/play/oFu2Qj2JC6LuGbcYfz8jK1hbY-wGHysPBB1CNJo258SAtBKxc8GMK8Z8gbIy3Oi2_jvAn0EEbXD5F4R3.L3JBr3BHdKD6w6pS?autoplay=true&startTime=1740489296000

B1: AI for Medicine and Healthcare

https://zoom.us/rec/play/XQJMoV7A45ly7ggk8jVZHxcIzb1K4GkU6LSrtB4brwJHtlMyTw_QT5GpWaNAmZYyCmwaixlThXX4kRjk.7xy0OeOZokUB5JMr?autoplay=true&startTime=1740491090000

=> student talk and project update, milestone discussion

CoLab enterprise exercise materials

 

https://cloud.google.com/blog/products/ai-machine-learning/running-alphafold-on-vertexai?_gl=1*1jnafks*_ga*MTgyMTg3NzI4OS4xNzQxMjg5MTMy*_ga_WH2QY8WWF5*MTc0MTI4ODUwNC4yLjEuMTc0MTI4ODUyOS4zNS4wLjA.&e=48754805

github in colab directly

 

Quick correction:

# Replace USERNAME with your GitHub username

# Replace TOKEN with your Personal Access Token

import os

# Set your GitHub credentials

GITHUB_USERNAME = "YOUR_GITHUB_USERNAME"

GITHUB_TOKEN = "YOUR_GITHUB_TOKEN_HERE"

# Clone the private repo using HTTPS, including username and token

!git clone https://{GITHUB_USERNAME}:{GITHUB_TOKEN}@github.com/repo_name/team_123.git

math formula between transformer encoder and decoder

The transformer architecture splits into an encoder and a decoder, and while both use the scaled dot‐product attention mechanism, they differ in how and where this mechanism is applied.

---

Scaled Dot-Product Attention (Common to Both)

At the core of both components is the scaled dot‐product attention defined as:

$$\text{Attention}(Q, K, V) = \text{softmax}\!\left(\frac{QK^T}{\sqrt{d_k}}\right) V$$

where:

• $Q$ (queries),
• $K$ (keys),
• $V$ (values), and
• $d_k$ is the dimensionality of the keys.

---

Encoder

In the encoder, every layer performs self-attention on the input sequence. Here, the queries, keys, and values are all derived from the same input $x$:

$$Q = xW^Q,\quad K = xW^K,\quad V = xW^V$$

Thus, the encoder’s self-attention is computed as:

$$\text{SelfAttention}_{\text{enc}}(x) = \text{softmax}\!\left(\frac{(xW^Q)(xW^K)^T}{\sqrt{d_k}}\right)(xW^V)$$

This mechanism allows each token in the input to attend to all other tokens, integrating contextual information across the entire sequence.

---

Decoder

The decoder is more complex because it has to generate an output sequence while incorporating information from the encoder. It uses two main attention mechanisms:

1. Masked Self-Attention:

   The decoder first applies self-attention to its own previous outputs. To maintain the autoregressive property (i.e., ensuring a token only depends on earlier tokens), a mask $M$ is applied. This mask typically sets the upper triangular part of the attention matrix to a very negative value, so that the softmax zeroes out any attention weights corresponding to future tokens. Formally:

   $$\text{MaskedAttention}(Q, K, V) = \text{softmax}\!\left(\frac{QK^T}{\sqrt{d_k}} + M\right)V$$

   Here, $Q$, $K$, and $V$ are derived from the decoder’s own input (previously generated tokens).

2. Encoder-Decoder (Cross) Attention:

   After the masked self-attention, the decoder incorporates information from the encoder. In this step, the queries come from the decoder (from the output of the masked self-attention), while the keys and values come from the encoder’s final output. The formula is:

   $$\text{EncDecAttention}(Q_{\text{dec}}, K_{\text{enc}}, V_{\text{enc}}) = \text{softmax}\!\left(\frac{Q_{\text{dec}} K_{\text{enc}}^T}{\sqrt{d_k}}\right)V_{\text{enc}}$$

   This step allows the decoder to "look" at the input sequence and incorporate context from the encoder into the output generation.

---

Summary of Differences

• Input Sources:

  • Encoder: Uses self-attention where $Q$, $K$, and $V$ are all derived from the same input $x$.
  • Decoder: Uses two attention mechanisms: masked self-attention on its own output (with a mask $M$) and cross-attention that uses the encoder's outputs for $K$ and $V$.

• Masking:

  • Encoder: No masking is necessary; all tokens attend to each other.
  • Decoder: Uses a mask in the self-attention to prevent future tokens from being attended to, preserving the autoregressive property.

• Attention Layers:

  • Encoder: A single self-attention layer per encoder block.
  • Decoder: Two sequential attention layers (masked self-attention followed by encoder-decoder attention) in each decoder block.

These differences in the attention formulas are key to enabling the decoder to generate coherent output sequences while leveraging the complete context provided by the encoder. 

alphafold 3 on ODU wahab

 

• Load modules
• module load alphafold/3.0.1
• Request a high memory GPU node 
• salloc -p high-gpu-mem --gres gpu=1
• Run the software, replacing with your directories
•  crun.alphafold run_alphafold.py --output_dir=/home/tstil004/alphafold3/output --json_path=/home/tstil004/alphafold3/gmp.json --model-dir=/home/tstil004/alphafold3/models
• 
  

mit flow

 

Introduction to Flow Matching and Diffusion Models

MIT Computer Science Class 6.S184: Generative AI with Stochastic Differential Equations

https://diffusion.csail.mit.edu/

https://github.com/eje24/iap-diffusion-class/tree/main/docs

ODU HPC

 Since HPC is generally a type of computer that is familiar to many researchers, we offered "Intro to HPC" workshops to help people onboard to using the system. We just had offered the last one last week. We recorded the workshop—anyone is welcome to watch this to learn:

“Introduction to HPC -- 2025-02-25 Workshop raw recording”

https://odumedia.mediaspace.kaltura.com/media/1_to1ksub3

Length: 03hr 32m 45s

Slides: https://drive.google.com/file/d/1h4-E335c4YnaiClAO9N3A63PSq6YYqXo/view .

Our complete  documentation can be read at: https://wiki.hpc.odu.edu/ .

Biomedical Data Science: Mining and Modeling (Spring 2025) Yale,

 

https://cbb752b25.gersteinlab.org/syllabus

Syllabus

Please see last year’s syllabus (with slide packs at the bottom) for previews for this year’s lectures, which will be slightly different. For this year, an updated slide pack will be posted after the lecture. (If it is substantially different from ‘22, an updated video will also be posted.) Video recordings for this year’s lectures can be found in the Media Library tab in Canvas.

	Topic	2025 Spring’s Lecture	2023 Spring’s Lecture	Comment	Lecture Summary
1/13	*YALE* Spring term classes begin, 8.20 a.m.
1/13	Introduction	25i1-25i2a	23i1, 23i2a
1/15	DATA - Proteomics I	25d3	23d3	Suggested Reading	25d3
1/22	DATA - Proteomics II	25d4	23d4	Suggested Reading	25d4
1/27	DATA - Genomics I	25d1	23d1		25d1
1/29	DATA - Genomics II	25d2	23d2		25d2
2/3	DATA - Knowledge Representation & Databases	25d5	23d5		25d5
2/5	MINING - Personal Genomes + Seq. Comparison	25i2b,25m3	23i2b,23m3
2/10	MINING - Seq. Comparison (con’t) + Multi-seq Alignment + Fast Alignment	25m3, 25m4, 25m5	23m3, 23m4, 23m5		2512b,25m3-part1
2/12	MINING - Variant Calling (incl. a focused section on SVs)+ Basic Multi-Omics	25m6a，23m6b, 23m7-pt1	23m6a, 23m6b,23m7
2/17	Quiz on 1st Half	quiz1 study guide
2/19	MINING - Supervised Mining #2 + Deep Learning Fundamentals #1		23m8b,23m8c
2/24	MINING - Deep Learning Fundamentals #2 + Unsupervised Mining #1		23m9a,23m9c
2/26	MINING - Unsupervised Mining #2 + Single-Cell Analysis #1		23m9d, 23m9e
3/3	MINING - Single-Cell Analysis #2 + Biomedical Image Analysis		23t1
3/5	MINING - Network Analysis		23m10a, 23m10b, 23m10c, 23m10d, 23m10e
3/7	Spring break begins
3/24	MINING - Privacy		23m11a, 23m11b
3/26	MINING/MODELING - Deep Learning Advanced I		23m12a
3/31	MINING/MODELING - Deep Learning Advanced II		23m12b
4/2	SIMULATION - Protein Simulation I		23s1
4/7	SIMULATION - Protein Simulation II		23s2
4/9	SIMULATION - Protein Simulation III		23s3
4/14	SIMULATION - Protein Simulation IV		23s4
4/16	SIMULATION - Protein Simulation V		23s5
4/21	Quiz on 2nd Half
4/23	Final Presentations
4/25	*YALE* Classes end; Reading period begins
5/1	*YALE* Final examinations begin
5/7	*YALE* Final examinations end