FlowingFlowing
← Back to all development cases
This case was judged by an anonymized AI reviewer and has not been independently reviewed by people; inspect the draft, cards, and both outputs to assess it.
AutocompleteWith cards better6 / 280 · beu-89bf16c0a707dfcf

A Review of BioTree Construction in the Context of Information Fusion: Priors, Methods, Applications and Trends

Quantitative biology · 2410.04815v2

FLOWING EVIDENCE BENCHMARK

How do we compare writing with and without cards?

At the same writing position, the same model produces one first result with cards and one without. All 280 pairs enter the overall anonymized review, which assesses whether a decision not to continue fits the draft.

Overall: 138 with-card wins, 26 ties, 107 no-card wins, 5 pairs where neither was usable, and 4 inconclusive judgments. Both sides wrote text in 218 pairs; 48 had a deliberate stop and 14 included an invalid output.

SAME DRAFT · WITH OR WITHOUT CARDS

01 · FIXED WRITING POSITION

MANUSCRIPT

Same writing position ▌

Matched autocomplete at the same draft position

02 · TWO MATCHED VERSIONS

SHARED BY BOTH

Draft, writing task, and generation model

A · With cards

Relevant literature cards supplied

B · Without cards

No literature cards supplied

Generation model

Same model and version

A → first output

B → first output

Anonymized AI judge

First outputs are anonymized as X and Y

Continuation review: accuracy, fit to the writing task and usability
Returns: X preferred / tie / Y preferred / both unusable

Review again with the answer order swapped

HOW DOES BLIND JUDGING WORK?

① Anonymize both outputs
The AI judge sees the same draft and both first outputs without knowing which received cards.

② Compare and swap order
The judge applies task-specific criteria in X/Y and then Y/X order.

③ Review disagreements
A third pass resolves disagreements. Incomplete reviews remain in the denominator.

How are positions and cards selected?

We selected 40 writing positions in each of seven fields, for 280 comparisons. The position and relevant literature cards were set before each comparison, so both versions began from the same draft.

The main result retains every frozen position, including places where cards might not add a useful next move. This avoids selecting cases based on their outputs, but 49% describes the whole cohort, not a win rate where evidence clearly had an opportunity to help. This cohort did not record that opportunity label for each position before generation, so those two groups cannot yet be separated reliably.

CARD USE · 140/280

Reported card use in continuations with text

The model reports whether a card affected its continuation. In 140 of 280 tests it produced continuation text and reported card use. The previous count of 150 included 10 cases without continuation text and has been corrected.

COMPARISON · 70 / 7 / 30

Which written version was better?

Of the 107 reported-use cases where both versions produced text, the blinded AI judge preferred the with-card version 70 times, found 7 about the same, and preferred the without-card version 30 times.

How are the two table percentages calculated?

ADOPTION COVERAGE

140reported use with text
÷
280all positions
=
50%

WITH-CARD VERSION PREFERRED

70with cards better
÷
107comparable tests
=
65%

The percentages use different groups: 50% counts continuations with text reporting card use among all 280 tests; 65% covers the 107 such cases with text on both sides. The generating model reports card use, while an anonymized AI judge compares writing quality.

Card use is self-reported by the generator. The anonymized AI judge sees the draft and source snippets, so source-aware preference remains possible; there was no independent human review. The overall result describes these 280 frozen positions, not a win probability for any writing task.

Manuscript writing position

Text excerpt · not a PDF page

Research manuscript · excerpt

A Review of BioTree Construction in the Context of Information Fusion: Priors, Methods, Applications and Trends

9 Current Limitations of BioTree Construction > 9.2 Challenges of Deep Learning-Based BioTree Construction Methods

… gene transfer or genome duplications are challenging to capture within traditional frameworks [ 231 ] . Bayesian approaches, despite offering flexibility through priors, are highly sensitive to model selection, where incorrect assumptions can lead to biased or erroneous results. For example, in polyploid plants or recombinant pathogens with intricate evolutionary histories, classical models often fail to provide biologically plausible insights. Incorporating information fusion techniques that blend empirical data with adaptive model selection may offer a promising avenue to address this gap.

…evolutionary trajectories. However, classical BioTree construction methods are predominantly designed for single-data-type analysis and lack robust mechanisms for combining multiple data sources [ 208 ] . When evolutionary signals conflict across omics layers, these methods fail to produce reliable integrated results. This deficiency hampers the holistic understanding of evolutionary processes and multi-level biological systems. While deep learning approaches have begun to leverage data-driven strategies for fusion, classical methods remain inadequate in addressing this integration challenge.

↓ Next paragraph was generated here

The highlighted text is near the writing position. This is a layout of the evaluation input, not a PDF screenshot.

A matched comparison at the same writing positionOverall AI review: With cards better

Both arms use the same model, task and manuscript context. Only the retrieved passages differ. The first recorded output from each arm is shown below.

Read overall AI review rationale

A = without cards · B = with cards

Both outputs attempt to continue the paragraph by identifying challenges with deep learning-based BioTree construction methods, which is a logical next move after noting classical methods are inadequate and deep learning has 'begun to leverage data-driven strategies.' However, Output A contains an unsupported concrete claim: it states 'curated phylogenetic datasets with consistent annotation across omics layers remain scarce, and the scarcity of labeled evolutionary events constrains supervised training.' This is a specific empirical claim about dataset scarcity and supervised training constraints for phylogenetic/omics data that is not supported by any of the provided sources (which discuss multi-omics integration, cancer classification, EEG, and cellular imaging, but not phylogenetic dataset curation or labeled evolutionary events). Output B makes more general, defensible claims about data quantity/quality dependence and interpretability challenges of complex nonlinear architectures—claims well-aligned with the sources' emphasis on deep learning as 'black boxes' lacking transparency (e.g., the Explainable AI survey explicitly notes DL models 'often function as black boxes, lacking transparency in decision-making processes'). Output B's focus on interpretability limitations directly connects to the manuscript's established theme of needing robust mechanisms and holistic understanding, without introducing unsupported specifics. Both are insertable without repair, but Output A's unsupported empirical claim about phylogenetic dataset scarcity reduces its utility.

Evidence

Same model and position · retrieved passages supplied · first output

Preferred by blind review

Deep learning-based BioTree construction methods face several unresolved challenges. Their performance is tightly coupled to the quantity and quality of training data, and the complex, nonlinear architectures that enable multi-omics fusion also render the resulting trees difficult to interpret, limiting their biological utility.

Passages supplied to the Evidence version

Card 1 · Explainable AI Methods for Multi-Omics Analysis: A Survey

Advancements in high-throughput technologies have led to a shift from traditional hypothesis-driven methodologies to data-driven approaches. Multi-omics refers to the integrative analysis of data derived from multiple 'omes', such as genomics, proteomics, transcriptomics, metabolomics, and microbiom…

Read full passage excerpt

Advancements in high-throughput technologies have led to a shift from traditional hypothesis-driven methodologies to data-driven approaches. Multi-omics refers to the integrative analysis of data derived from multiple 'omes', such as genomics, proteomics, transcriptomics, metabolomics, and microbiomics. This approach enables a comprehensive understanding of biological systems by capturing different layers of biological information. Deep learning methods are increasingly utilized to integrate multi-omics data, offering insights into molecular interactions and enhancing research into complex diseases. However, these models, with their numerous interconnected layers and nonlinear relationships, often function as black boxes, lacking transparency in decision-making processes. To overcome this challenge, explainable artificial intelligence (xAI) methods are crucial for creating transparent models that allow clinicians to interpret and work with complex data more effectively. This review exp …

Card 2 · Supervised multiple kernel learning approaches for multi-omics data integration

BACKGROUND: Advances in high-throughput technologies have originated an ever-increasing availability of omics datasets. The integration of multiple heterogeneous data sources is currently an issue for biology and bioinformatics. Multiple kernel learning (MKL) has shown to be a flexible and valid app…

Read full passage excerpt

BACKGROUND: Advances in high-throughput technologies have originated an ever-increasing availability of omics datasets. The integration of multiple heterogeneous data sources is currently an issue for biology and bioinformatics. Multiple kernel learning (MKL) has shown to be a flexible and valid approach to consider the diverse nature of multi-omics inputs, despite being an underused tool in genomic data mining. RESULTS: We provide novel MKL approaches based on different kernel fusion strategies. To learn from the meta-kernel of input kernels, we adapted unsupervised integration algorithms for supervised tasks with support vector machines. We also tested deep learning architectures for kernel fusion and classification. The results show that MKL-based models can outperform more complex, state-of-the-art, supervised multi-omics integrative approaches. CONCLUSION: Multiple kernel learning offers a natural framework for predictive models in multi-omics data. It proved to provide a fast and …

Card 3 · LASSO–MOGAT: a multi-omics graph attention framework for cancer classification

The application of machine learning (ML) methods to analyze changes in gene expression patterns has recently emerged as a powerful approach in cancer research, enhancing our understanding of the molecular mechanisms underpinning cancer development and progression. Combining gene expression data with…

Read full passage excerpt

The application of machine learning (ML) methods to analyze changes in gene expression patterns has recently emerged as a powerful approach in cancer research, enhancing our understanding of the molecular mechanisms underpinning cancer development and progression. Combining gene expression data with other types of omics data has been reported by numerous works to improve cancer classification outcomes. Despite these advances, effectively integrating high-dimensional multi-omics data and capturing the complex relationships across different biological layers remain challenging. This article introduces Least Absolute Shrinkage and Selection Operator–Multi-omics Gated Attention (LASSO–MOGAT), a novel graph-based deep learning framework that integrates messenger RNA, microRNA, and DNA methylation data to classify 31 cancer types. By utilizing differential expression analysis (DEG) with Linear Models for Microarray (LIMMA) and LASSO regression for feature selection and leveraging graph atten …

Card 4 · Deep learning-based electroencephalography analysis: a systematic review

CONTEXT: Electroencephalography (EEG) is a complex signal and can require several years of training, as well as advanced signal processing and feature extraction methodologies to be correctly interpreted. Recently, deep learning (DL) has shown great promise in helping make sense of EEG signals due t…

Read full passage excerpt

CONTEXT: Electroencephalography (EEG) is a complex signal and can require several years of training, as well as advanced signal processing and feature extraction methodologies to be correctly interpreted. Recently, deep learning (DL) has shown great promise in helping make sense of EEG signals due to its capacity to learn good feature representations from raw data. Whether DL truly presents advantages as compared to more traditional EEG processing approaches, however, remains an open question. OBJECTIVE: In this work, we review 154 papers that apply DL to EEG, published between January 2010 and July 2018, and spanning different application domains such as epilepsy, sleep, brain-computer interfacing, and cognitive and affective monitoring. We extract trends and highlight interesting approaches from this large body of literature in order to inform future research and formulate recommendations. METHODS: Major databases spanning the fields of science and engineering were queried to identif …

Card 5 · Multi-omics Prediction from High-content Cellular Imaging with Deep Learning

High-content cellular imaging, transcriptomics, and proteomics data provide rich and complementary views on the molecular layers of biology that influence cellular states and function. However, the biological determinants through which changes in multi-omics measurements influence cellular morpholog…

Read full passage excerpt

High-content cellular imaging, transcriptomics, and proteomics data provide rich and complementary views on the molecular layers of biology that influence cellular states and function. However, the biological determinants through which changes in multi-omics measurements influence cellular morphology have not yet been systematically explored, and the degree to which cell imaging could potentially enable the prediction of multi-omics directly from cell imaging data is therefore currently unclear. Here, we address the question of whether it is possible to predict bulk multi-omics measurements directly from cell images using Image2Omics - a deep learning approach that predicts multi-omics in a cell population directly from high-content images of cells stained with multiplexed fluorescent dyes. We perform an experimental evaluation in gene-edited macrophages derived from human induced pluripotent stem cells (hiPSC) under multiple stimulation conditions and demonstrate that Image2Omics achi …

Model-reported card use

Card 1 · fact

Affected output: the complex, nonlinear architectures that enable multi-omics fusion also render the resulting trees difficult to interpret

Model explanation: The row's proposition that deep learning models with numerous interconnected nonlinear layers function as black boxes supplied the factual basis for the interpretability challenge stated in the generated text.

Card 4 · direction

Affected output: Deep learning-based BioTree construction methods face several unresolved challenges.

Model explanation: This row's framing of deep learning advantages as an open question guided the choice to open the paragraph with a general statement of unresolved challenges rather than a specific technical limitation.

No Evidence

Same model and position · no retrieved passages · first output

Not preferred in this comparison

Deep learning-based BioTree construction methods face several challenges that limit their reliability and broader adoption. These models typically require large-scale, high-quality training data, yet curated phylogenetic datasets with consistent annotation across omics layers remain scarce, and the scarcity of labeled evolutionary events constrains supervised training.

Earlier card-use comparison

The earlier diagnostic compared only cases with reported card use and text on both sides; its scope differs from the overall AI review above.

The blinded model review preferred the Evidence version. Its original rationale appears below.

Read the original model judgment

Both outputs attempt to continue the paragraph about challenges in BioTree construction methods. Output A makes two grounded claims: (1) deep learning performance depends on data quantity/quality (supported by the general context about data-driven approaches and the EEG review's observation about data variability), and (2) complex nonlinear architectures make trees difficult to interpret (supported by the Explainable AI survey's discussion of deep learning models as 'black boxes' lacking transparency). Output B adds unsupported material claims: 'curated phylogenetic datasets with consistent annotation across omics layers remain scarce' and 'scarcity of labeled evolutionary events constrains supervised training' — neither the draft nor any source mentions phylogenetic datasets, consistent annotation across omics layers, labeled evolutionary events, or supervised training constraints in this context. These are fabricated specifics about dataset availability and training paradigms that have no basis in the provided materials. Output A stays closer to general, defensible limitations that align with source themes about interpretability and data dependence without inventing specific dataset characteristics.

The passages are input material; their presence does not mean an output used them or that they support every claim. Draft and source passages are excerpted; no simulated scores or PDF appear here.