First Doctoral Thesis in our Lab
Bowen F., a doctoral student from our BMDS lab, successfully graduated. He is the first doctoral student to graduate from our lab. We celebrated his achievement with thoughtful gifts along with smiles, hugs, an aperitif, friends, crossfit, and fun.
His Ph.D. defense took place in the new building of the ETH-Department of Biosystems Science and Engineering (D-BSSE) on the Schällemätteli campus in Basel.
Bowen's main contributions were the article on 1) “Prediction of recovery from multiple organ dysfunction syndrome (MODS) in pediatric sepsis patients” external page Bowen F., et al. 2022 and 2) the manuscript in preparation about Modelling PCR bias in DNA sequence. He also contributed to other papers seen in the comments.
Children with sepsis often experience rapid progression or recovery within days of admission, and the persistence of Multiple Organ Dysfunction Syndrome (MODS) in the 1st week is linked to high mortality and morbidity. A machine learning-based model was developed to forecast MODS recovery to zero or single organ dysfunction a week in advance in the Swiss Pediatric Sepsis Study (SPSS) cohort. Notably, cardiovascular and respiratory systems played critical roles in early recovery prediction, with top-ranking variables associated with these systems. This underscores the need for tailored care for MODS patients with cardiovascular or respiratory failure. This pioneering study, validated across continents from Switzerland to the USA, highlights the potential of integrating the model into electronic health records for enhanced patient assessment and triage in pediatric sepsis.
In biotech research, polymerase chain reaction (PCR) bias - or DNA amplification - is a big problem because it causes differences in how well DNA is amplified. Given the exponential nature of PCR, even small variations in efficiency during PCR can lead to big differences in the results. To tackle this issue, Bowen and colleagues suggested a new computational pipeline. This method uses DNA sequencing data to study PCR bias in detail. It starts with a statistical model that looks at every step of the sequencing process to estimate how well PCR is working for each DNA sequence. Then, a special deep learning model is used to predict PCR efficiency based on the structure of the DNA sequence. Finally, a complex method is used to find patterns - or so called motifs - in the data that are linked to lower PCR efficiency. These findings could make PCR more reliable, which would be helpful for many applications like studying gene expression, profiling microbial communities, DNA storage, and more.
We congratulate Bowen once again and wish him the best as he completes his 2nd contribution and extend our best wishes for his future career endeavors. We are confident that his datas cience and fitness dedication combined with his talent will lead to continued success in his professional and life journey. 恭喜
Visit our external page LinkedIn BMDS for more news.