K-means clustering details:

Cluster	Members
Cluster 1	SP31, SP32, SP33, SP41, SP42, SP43
Cluster 2	SP1, SP2, SP3
Cluster 3	SP21, SP22, SP23, SP51, SP52, SP53

Clustering of samples based on similarity of concentration of PAHs

Principle Component Analysis: Sample 1 and Sample 4 share similarity in low concentration of various PAHs while Sample 2 and Sample 5 share similarity in high concentration of various PAHs. Sample 3 has a more or less balanced concentration of PAHs.

Based on the two samples, sample 1 and sample 2, we scored PAHs that were highly concentrated in a relatively balanced distribution in both samples. These included Benz(a)anthracene, pyrene, acenaphthylene, fluoranthene, 1- methylnaphthalene and phenanthrene.

PAH ANALYSIS BASED ON SAMPLE SITES

Sample 1: Jinja Kaloli

The PAH levels in Jinja Kaloli are relatively low for most compounds, with only 2-Methylnaphthalene and 1-Methylnaphthalene showing significant concentrations. This suggests that this site might have minimal PAH pollution or contamination.

Sample 2: Bwaise II

The concentrations of PAHs in Bwaise II are notably high across several compounds. Almost all the PAHs listed have been detected in significant amounts, with 2-Methylnaphthalene, 1-Methylnaphthalene, and Fluorene having particularly high concentrations. Bwaise II seems to be a hotspot for PAH contamination. The presence of multiple PAHs in significant concentrations indicates potential industrial activities or other sources of pollution in the area.

Sample 3. Katanga

Katanga shows a mixed profile with some PAHs like Naphthalene, 2-Methylnaphthalene, and 1-Methylnaphthalene present in high concentrations, while others like acenaphthylene, Fluoranthene, and Pyrene are absent. The presence of specific PAHs and absence of others might indicate specific sources of pollution in Katanga. It's essential to investigate further to determine the exact cause of these PAH concentrations.

Sample 4: Kalerwe

Kalerwe has a unique profile with a very high concentration of Naphthalene. However, many other PAHs are absent. The high concentration of Naphthalene suggests a specific source of contamination in Kalerwe. It might be related to specific industrial activities or other localized sources of pollution.

Sample 5: Kikoni

Kikoni has high concentrations of several PAHs, notably Naphthalene, 2-Methylnaphthalene, 1-Methylnaphthalene, and Pyrene. This suggests a significant level of PAH contamination, possibly from various sources.

General Observation and Recommendations

The PAH analysis indicates varying levels of contamination across the different sample sites. Bwaise II and Kikoni stand out as significant hotspots for PAH contamination, while Jinja Kaloli seems to have minimal PAH pollution. It's crucial to understand the sources of these PAHs to implement effective mitigation measures.

SCORING RELEVANT PAHS

The concentrations of different PAH compounds vary across the water samples. This variability suggests that the sources of these compounds are diverse, and not all PAHs are equally present in the samples. Some PAH compounds are present at higher levels compared to others. This indicates that these compounds may originate from different sources, such as industrial processes, vehicular emissions, or natural occurrences.

SPECIFIC SAMPLES

Low Concentration Samples: Sample 1 and Sample 4 have the least concentration of PAHs. These samples may come from sources with lower levels of pollution or better filtration methods.

High Concentration Samples: Sample 2 and Sample 5 have the highest concentration of PAHs. These could be from sources with higher levels of pollution or less effective filtration methods.We therefore decided to focus on these two samples to choose ideal PAH candidates for the biosensor.

IDEAL CANDIDATES FOR THE BIOSENSOR

Pyrene: This compound scores highest in all ranking categories, making it very ideal for the biosensor model. The high limit of detection of 0.3 for Pyrene when using GC-MS suggests that it is particularly well-suited for applications where the target analytes are expected to be present in higher concentrations. While GC-MS is a powerful tool for qualitative and quantitative analysis, its higher limit of detection may not be ideal for detecting trace amounts of substances. However, this is where the biosensor model excels. It could be particularly useful for real-time, on-site monitoring of pyrene in environments where low-level detection is crucial, such as in water quality assessment or air pollution monitoring. The biosensor's capabilities can fill the gaps left by the limitations of GC-MS, offering a more comprehensive and nuanced approach to pyrene detection.

Phenanthrene: This is also a priority PAH and is fairly distributed across the five samples. Given its prevalence and importance, it is a suitable candidate for detection. Due to its prevalence, phenanthrene can serve as a marker or indicator compound for the presence of other, potentially more harmful, PAHs. Its widespread occurrence across samples suggests that it is easier to detect, making it a suitable candidate for calibration and quality control in analytical methods.

CONCLUSION

Based on the observed variations in PAH concentrations and their distribution across different samples, the biosensor will be designed to detect two key compounds: Phenanthrene and Pyrene. These compounds are not only prevalent but also likely represent different sources, making them ideal markers for monitoring water quality.

By focusing on these two compounds, the biosensor aims to provide a reliable and effective method for detecting harmful PAH levels in water used for various purposes, thereby contributing to better environmental management and public health.