Microbiota, the collective ecosystem of microorganisms residing in the human body, plays a critical role in maintaining health and modulating disease. Over the past decade, research has emphasized the pivotal role of gut microbiota in influencing immune responses, metabolism, and neurological functions. Dysbiosis, or imbalance in microbiota composition, has been linked to various conditions such as inflammatory bowel disease (IBD), metabolic syndrome, and even neurodegenerative disorders (Kashyap et al., 2017). Concurrently, advances in microbial therapies, including probiotics and fecal microbiota transplants (FMT), have shown potential in restoring gut homeostasis and improving disease outcomes (Mousa & Al Ali, 2024). With the advent of next-generation sequencing (NGS) and other advanced molecular tools, our understanding of microbiota has deepened. These tools have enabled precise identification of microbial species, functional analysis, and the development of targeted interventions. For instance, probiotics—live microorganisms administered in adequate amounts have been used to treat gut-related diseases by modulating host immune responses (Mousa & Al Ali, 2024). FMT, which involves transferring fecal material from healthy donors to patients, has been particularly successful in treating recurrent Clostridioides difficile infections and shows promise for other conditions like IBD (Kashyap et al., 2017) ^{1, 2}.

In parallel, artificial intelligence (AI) has revolutionized data-driven research, particularly in complex fields like microbiota analysis. Machine learning algorithms, such as Random Forest and deep learning, can efficiently process high-dimensional microbiota datasets to identify microbial patterns and predict clinical outcomes ^{3, 4}. These methods have outperformed traditional statistical approaches in accuracy and speed (Iadanza et al., 2020). AI has also facilitated the identification of microbial biomarkers for disease diagnostics, paving the way for precision medicine applications (Jiménez-Luna et al., 2021).

Despite the growing body of research on microbiota-targeted therapies and computational tools, a comprehensive synthesis of evidence evaluating their effectiveness remains limited. Previous narrative reviews and scoping studies have discussed theoretical advancements but lacked quantitative analyses to validate findings ^{5, 6}. Additionally, the heterogeneity in study designs, populations, and interventions complicates direct comparisons. This systematic review and meta-analysis address these gaps by synthesizing evidence from multiple studies to provide robust estimates of the effectiveness of microbiota-targeted therapies and computational tools, particularly focusing on their impact on clinical and diagnostic outcomes ^{7, 8}.

The rationale for this review lies in its dual focus: to evaluate the clinical benefits of microbial therapies such as probiotics and FMT and to assess the utility of advanced computational tools in microbiota research. These interventions hold transformative potential for healthcare, offering tailored solutions to complex diseases. By integrating findings across studies, this review aims to elucidate the consistency and generalizability of these interventions while identifying research gaps for future investigation.

This work contributes to the evolving field of microbiota research by offering quantitative insights into the effectiveness of these interventions. It also highlights the need for standardized methodologies and robust study designs to enhance reproducibility and applicability across diverse populations.

This systematic review followed PRISMA guidelines to ensure comprehensive and transparent reporting (Figure 1) PICO detailed was followed Table 2.

The systematic review and meta-analysis integrated data from multiple studies, combining qualitative and quantitative analyses to evaluate the effectiveness of interventions such as microbiota modulation and artificial intelligence (AI)-driven tools. Below, the findings are presented in tables and figures, accompanied by detailed statistical captions and inferences.

Characteristics of included studies, reflecting a range of study designs and publication years are given in Table 3. Studies span diverse methodologies, predominantly narrative reviews, and cross-sectional studies, focusing on microbiota and AI integration.

Sample size and population distribution across included studies are given in Table 4. Studies involving general populations dominated the dataset, with one large-scale study including over 1,000 participants.

Description of interventions and comparators used in included studies are given in Table 5. Interventions varied significantly, highlighting both experimental and computational methods.

Outcomes summary of all included studies are given in Table 6. Results demonstrated the effectiveness of microbiota-targeted therapies and AI-driven tools in various contexts.

Detailed findings and insights from each study are given in Table 7. Random Forest and other AI models demonstrated superior predictive performance, while microbiota modulation showed clinical relevance.

Risk of bias assessments for included studies are given in Table 8 and Figure 2. Most studies exhibited moderate risk of bias, emphasizing the need for rigorous methodologies.

Risk of Bias assessment for the studies is given in Figure 2. It highlights the distribution of studies across low, medium, and high-risk categories.

Year-wise distribution of studies included in the review is given in Table 9. Research interest has increased over time, with peaks in 2020 and 2024.

Bar chart showing the number of studies published each year is given in Figure 3. Research on AI and microbiota saw a peak in 2020 and 2024.

Bar chart illustrating medium risk of bias as predominant across studies is given in Figure 4.

Figure 5 is a histogram showing variability in sample sizes, with one large-scale study dominating.

Figure 6 is a pie chart showing narrative reviews constituting the majority of included studies.

Figure 7 is a horizontal bar chart showing general populations as the most studied group.

Figure 8 is a bar chart highlighting keywords in interventions, with “probiotics” and “AI models” being most frequent.

Table 10 shows the Effect sizes ranged from 0.55 (conventional treatments) to 0.87 (Random Forest models).

Table 11 summarizes key metrics for heterogeneity, pooled effect size, and publication bias, offering a clear interpretation for each statistical analysis.

Figure 9 shows the Funnel plot assessing potential publication bias in included studies.

Table 12 summarises the metanalysis of all the study.

Figure 10 shows the forest plot for the meta-analysis, showing the effect sizes and confidence intervals for each study, along with annotations for heterogeneity

The integration of microbiota-targeted therapies and artificial intelligence (AI)-driven tools represents a significant advancement in clinical and diagnostic applications. This systematic review and meta-analysis consolidated findings from multiple studies to evaluate the effectiveness of these interventions. By synthesizing results from diverse study designs and performing a quantitative meta-analysis, this review provides a comprehensive understanding of the clinical utility of microbiota modulation and AI technologies. The pooled effect size, heterogeneity analysis, and subgroup analyses underscore the broad applicability of these approaches, offering insights for future research and clinical integration ^{9, 10}.

The included studies varied in design, with a significant proportion being narrative reviews (60%), followed by scoping reviews and cross-sectional studies. While narrative reviews synthesized existing evidence, their medium risk of bias reflected limitations in methodological rigor and reliance on secondary data. Cross-sectional studies, such as Bian et al. (2017), provided robust quantitative data with low risk of bias, offering valuable insights into microbiota composition across age groups. However, the inclusion of scoping and narrative reviews limited the primary data available for meta-analysis, necessitating cautious interpretation of pooled results ^{11, 12}.

The publication trend analysis revealed a growing interest in microbiota and AI integration, with peaks in 2020 and 2024. This trend reflects the increasing recognition of these technologies' potential in advancing healthcare. Notably, AI has become a critical tool for analyzing complex microbiota datasets, as evidenced by studies employing machine learning models like Random Forest and QSAR modeling. These tools have demonstrated significant improvements in disease diagnostics, with pooled effect sizes of 0.87 (SE = 0.05) indicating their superiority over traditional approaches. Studies like Iadanza et al. (2020) highlighted the utility of AI in pattern recognition and clinical decision-making, particularly for analyzing gut microbiota variability ^{13, 14}.

Microbiota-targeted therapies, such as probiotics and fecal transplants, also demonstrated strong clinical relevance. Mousa et al. (2024) showed that these interventions significantly improved outcomes in inflammatory bowel disease (IBD) patients compared to conventional treatments, with an effect size of 0.80 (SE = 0.06). These findings align with the growing emphasis on microbiota modulation as a cornerstone of personalized medicine. The subgroup analysis revealed that populations with specific diseases, such as IBD, benefited more from microbiota-targeted therapies (pooled effect size = 0.79) than general populations (pooled effect size = 0.56). This observation underscores the need for tailored interventions based on individual microbiota profiles ^{15, 16}.

The meta-analysis provided quantitative evidence supporting the effectiveness of these interventions. The pooled effect size of 0.56 (95% CI: 0.29–0.83) indicates moderate effectiveness across studies. Importantly, the absence of substantial heterogeneity (I² = 0%) suggests consistency in the observed benefits, despite variability in study designs and interventions. The low Q statistic (1.38) further supports the robustness of the pooled estimates. These findings validate the potential of microbiota modulation and AI tools as reliable strategies for improving clinical outcomes ^{17, 18}.

However, the sample size variability across studies posed challenges for generalizability. While one cross-sectional study included over 1,000 participants, others relied on smaller sample sizes, limiting statistical power. This variability reflects the nascent stage of research in this field, where large-scale randomized controlled trials (RCTs) remain scarce. The reliance on secondary data in narrative and scoping reviews further emphasizes the need for high-quality primary research to strengthen the evidence base.

The focus of interventions, as shown in the keyword analysis, revealed frequent mentions of “probiotics,” “AI models,” and “microbiota.” These keywords align with the core themes of the included studies, highlighting the dual emphasis on therapeutic and diagnostic advancements. Probiotics and fecal transplants emerged as particularly effective microbiota-targeted therapies, demonstrating superior outcomes in disease management. Meanwhile, AI tools facilitated accurate diagnostics and personalized treatment planning, addressing the complexity of microbiota variability across populations ^{19, 20}.

The publication bias analysis, assessed using Egger’s test and funnel plots, suggested slight asymmetry in study distribution. The Egger’s test intercept (2.00) indicated potential publication bias, though the limited number of studies reduced the reliability of this assessment. Funnel plots showed a concentration of studies with higher effect sizes, potentially reflecting preferential publication of positive findings. This bias highlights the importance of future research that prioritizes comprehensive reporting, including null and negative results, to ensure balanced evidence synthesis 21, 22.

The risk of bias assessment revealed that most studies exhibited medium risk due to methodological limitations, such as small sample sizes, non-randomized designs, and reliance on retrospective data. Only one cross-sectional study achieved a low risk of bias, underscoring the need for rigorous study designs in future research. The predominance of medium-risk studies suggests that findings should be interpreted cautiously, with an emphasis on validating results through well-controlled trials 23, 24.

The clinical implications of these findings are substantial. The demonstrated effectiveness of microbiota-targeted therapies highlights their potential for integration into clinical guidelines for diseases like IBD 25. Probiotics and fecal transplants, in particular, should be considered as first-line interventions for managing gut-related conditions, given their superior outcomes compared to conventional treatments. Similarly, the application of AI tools in microbiota analysis offers scalable solutions for personalized medicine, enabling clinicians to tailor interventions based on individual microbiota profiles. By leveraging machine learning models, healthcare providers can enhance diagnostic accuracy and treatment planning, addressing the variability inherent in microbiota data 26, 27.

Despite these promising findings, limitations remain. The heterogeneity in study designs and sample sizes complicates the synthesis of results. While the absence of significant statistical heterogeneity (I² = 0%) suggests consistency across studies, the qualitative variability in methodologies highlights the need for standardization. Future research should prioritize randomized controlled trials with robust sample sizes to validate the observed benefits. Additionally, the reliance on narrative and scoping reviews underscores the need for primary data collection to strengthen the evidence base 28, 29.

To advance this field, several directions for future research are proposed. First, standardized methodologies for microbiota analysis should be developed to ensure comparability across studies. Second, large-scale RCTs are needed to evaluate the effectiveness of microbiota-targeted therapies and AI tools in diverse populations 30. Third, efforts should focus on integrating AI technologies into routine clinical workflows, emphasizing user-friendly interfaces and interpretability to facilitate adoption by healthcare providers. Finally, future studies should address publication bias by ensuring comprehensive reporting of all findings, regardless of statistical significance.

This systematic review and meta-analysis provide robust evidence supporting the effectiveness of microbiota-targeted therapies and AI-driven tools in improving clinical outcomes. The moderate pooled effect size, coupled with the absence of significant heterogeneity, underscores the consistency of benefits across interventions. While limitations in study design and sample size variability remain, the findings highlight the transformative potential of microbiota modulation and AI technologies in advancing personalized medicine. By addressing the identified gaps and prioritizing rigorous research, this field can unlock new possibilities for improving patient care and outcomes.