Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

POMP: Pathology-omics Multimodal Pre-training Framework for Cancer Survival Prediction

Authors: Suixue Wang, Shilin Zhang, Huiyuan Lai, Weiliang Huo, Qingchen Zhang

IJCAI 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct extensive experiments on six cancer datasets from the Cancer Genome Atlas (TCGA) [Tomczak et al., 2015]. The experimental results show that POMP consistently outperforms existing state-of-the-art methods across six datasets.
Researcher Affiliation Academia 1 School of Information and Communication Engineering, Hainan University 2 College of Intelligence and Computing, Tianjin University 3 University of Groningen 4 School of Computer Science and Technology, Hainan University EMAIL, zhang shilin EMAIL, EMAIL
Pseudocode No The paper describes the methodology using textual explanations and a block diagram (Figure 1), but it does not contain explicitly labeled pseudocode or algorithm blocks.
Open Source Code Yes The code is available at https://github.com/Suixue Wang/POMP.
Open Datasets Yes We perform experiments using six cancer datasets obtained from the Cancer Genome Atlas (TCGA) data portal.
Dataset Splits Yes We assess all investigated methods using the same 5-fold cross-validation splits on each cancer dataset.
Hardware Specification Yes We implement our framework POMP using Py Torch and train it on 3 NVIDIA A100 GPUs.
Software Dependencies No The paper mentions using Py Torch for implementation and refers to the 'pydeseq2 package' and 'random survival forest (RSF)' but does not provide specific version numbers for any of these software components.
Experiment Setup Yes POMP is trained for 500 epochs during pretraining and 80 epochs during fine-tuning. For both training phases, we use Adam optimization with a weight decay of 1e2 and a learning rate of 5e-4. Since the pathological images have different sizes and are cropped into various sub-region numbers, we use a batch size of 1 with 50 forward accumulation steps (i.e., the actual batch size is equivalent to 50), then calculate the loss function and update the weights once.