International Society for Computational Biology - ISCB

06/19/2026

⌛️ Don't forget to book your hotel room TODAY to get the discounted room rate!

⏰ Only 2 days left for discounted hotel rates!

The deadline to book your room at the Washington Hilton in the official ISMB 2026 hotel block is this Friday, June 19 at 11:59pm ET. Rooms are filling up quickly, so if you haven't booked yet, don't wait!

Book your room using the personalized link in your registration confirmation email.

➡️ Why book at the conference hotel?
Staying on-site puts you seconds away from the conference, with sessions, networking, and your accommodation all in one place.

Plus, everyone booking in the official block will be entered into drawings for:
🔹The grand prize of a complimentary registration to a future ISMB conference
🔹Daily lunch gift card drawings during the conference

For more venue details, click here: https://www.iscb.org/ismb2026/general-info/venue-accommodation-and-travel

06/19/2026

📣 Last call for submissions to the YBS 2026 Student Challenge!

This year's theme is "Bioinformatics to Understand Mental Health Disorders."

If you're ready to join:
1️⃣ Head to https://www.iscb.org/ybs2026/register
2️⃣ Click the blue registration button
3️⃣ Add your project title and abstract during registration!

📅 Deadline to register for the challenge: Sunday, June 21, 2026

Full details on the YBS Student Challenge can be found here: https://www.iscb.org/ybs2026/whats-happening/student-challenge

📌If you're already registered, you can modify your registration to add your project details using the link in your confirmation email.

06/19/2026

Spurious proteins, sequences that arise from gene prediction errors rather than real biology, pose a growing challenge as protein databases expand. This study tests whether AlphaFold2, AlphaFold3, and ESMFold can distinguish spurious sequences from real proteins using structure prediction confidence scores. All three methods assign unexpectedly high confidence to short spurious sequences, but separation from real proteins improves meaningfully for sequences longer than 100 amino acids.

Read the full paper at https://doi.org/10.1093/bioadv/vbag160

06/19/2026

Most spatial omics tools are built for transcriptomics and struggle with the distinct challenges of mass spectrometry data, including missing values, sparse sampling, and the need to integrate multiple molecular types from the same tissue. spammR (Spatial Analysis of Multi-Omics Measurements in R) addresses this gap with an end-to-end analysis pipeline covering data ingestion, imputation, feature selection, pathway enrichment, and visualization, all built on the Bioconductor SpatialExperiment framework for broad interoperability. The package was validated across pancreatic cancer proteomics, rat brain lipidomics and proteomics integration, and continental-scale soil metagenomics from the 1000 Soils project, demonstrating its flexibility across both tissue and geographic distance scales.

Find the full paper at https://doi.org/10.1093/bioadv/vbag163

06/18/2026

🚨 There's only 1 week left to register for tutorials at ISMB 2026

Tutorials are your chance to get hands-on training directly from leaders in a range of established and emerging topics in computational biology and bioinformatics. 🧬💻

Tutorial registration deadline: Thursday, June 25, 2026

Check out the full list of tutorials here: https://www.iscb.org/ismb2026/whats-happening/tutorials

💻 Already registered for ISMB 2026?
You can still add tutorials to your existing registration:
1️⃣ Find your ISMB registration confirmation email
2️⃣ Click the link to modify your registration
3️⃣Select the "Modify Registration" button on the confirmation page
4️⃣Head to the Tutorials page and add the tutorials you'd like to attend
5️⃣Click the continue buttons to complete your updated registration, including paying your new outstanding balance

06/18/2026

Integrating diverse cancer data types, such as RNA expression, tumor images, somatic mutations, and clinical notes, typically requires computationally expensive coordination between AI models. This study challenges that assumption by showing that simply pooling embeddings from unaligned, independently trained models still preserves meaningful biological signals. Deep learning models trained on these pooled vectors successfully identified both data modality composition and tumor of origin across breast, bladder, lung, thyroid, colon, and stomach cancers.

Read the full paper at https://doi.org/10.1093/bioadv/vbag159

06/18/2026

Detecting rare vertebrates in sedimentary ancient DNA (sedaDNA) is a major challenge, as these organisms represent a tiny fraction of metagenomic datasets. This study benchmarks a stringent bioinformatic pipeline using both synthetic and empirical sedaDNA data from IODP Expedition 382 in the Scotia Sea, testing taxonomic assignment precision across varying fragment quantities, DNA damage levels, and sediment ages. Results establish that at least 250 and 500 DNA fragments are needed for reliable family- and genus-level identification, respectively.

Find the full paper at https://doi.org/10.1093/bioadv/vbag113

06/17/2026

⏰ Only 2 days left for discounted hotel rates!

The deadline to book your room at the Washington Hilton in the official ISMB 2026 hotel block is this Friday, June 19 at 11:59pm ET. Rooms are filling up quickly, so if you haven't booked yet, don't wait!

Book your room using the personalized link in your registration confirmation email.

➡️ Why book at the conference hotel?
Staying on-site puts you seconds away from the conference, with sessions, networking, and your accommodation all in one place.

Plus, everyone booking in the official block will be entered into drawings for:
🔹The grand prize of a complimentary registration to a future ISMB conference
🔹Daily lunch gift card drawings during the conference

For more venue details, click here: https://www.iscb.org/ismb2026/general-info/venue-accommodation-and-travel

06/17/2026

A key innovation in QProMS (Quantitative Proteomics Made Simple) is mixed imputation, a new framework for handling missing values that distinguishes between values missing at random and those missing due to low protein abundance. Benchmarking shows this approach outperforms the widely used downshift method and closely matches the more computationally expensive missForest, with particular advantages in affinity purification and proximity labeling experiments. All analyses are exportable as shareable, reproducible reports.

Explore the full study at https://doi.org/10.1093/bioadv/vbag158

06/17/2026

MOBAA is a new R package for identifying biologically distinct population subgroups from multi-omic data without relying on parametric distributional assumptions. The tool combines ensemble biclustering across individual omic layers with Jaccard-based hierarchical clustering to form cross-omic modules, then uses permutation-derived empirical p-values to assess their significance. Applied to TCGA kidney cancer and breast cancer cohorts, MOBAA recovered established disease subtypes with high cluster purity and revealed subtype-specific pathway signatures.

See the full study at https://doi.org/10.1093/bioadv/vbag156