Cancer Genome Sequencing

We performed whole-genome-sequencing and amplicon-sequencing on tumor tissue. Our goal was to see, if our conceptualized pipeline was practical and how long the process takes, since time is crucial resource for our intraoperative approach. We used the nanopore sequencing technology to get the raw data of the cancer tissue, due to its advantages helping us overcome some of the problems we identified in the conceptualization of our project. This nanopore sequencing approach does have some drawbacks, but we believe that the advantages heavily outweigh the disadvantages.


Advantages over other sequencing technologies

One big advantage is the fact that nanopore sequencing provides real-time results, as in the data is continuously streamed out while the machine is running, opposed to other sequencing technologies, whose results are only provided at the end of their runtime. This means that we can take data points at any time and work with them, as well as take data points from a later time to see how the read quality and the coverage improves. Another advantage is, that nanopore-sequencing is faster than other sequencing methods. As mentioned above, time is one of the main hurdles to overcome, which means speed is key. It is capable of sequencing 50 – 250 bases per second per pore and a four-color fluorophore system will sequence over 500 bases per second [1]. The workflows are kept simple by avoiding amplification or light-based measurement steps [2]. The coverage plots after six hours and one day of our sequencing run can be seen in Figure 1 below.



For our analysis, we used the GridION model provided to us by the CeBiTec, which uses up to five MinION flow cells. For later use in the real world, it would make sense to use the PromethION model, since this variant is designed to use up to 48 flow cells [2]. A full PromethION setup with 48 flow cells should lead to a noticeable time gain, which would further improve the workflow.


Data processing workflow

Figure 2: The workflow that was used to process the raw data provided by the nanopore sequencer.

The figure shows the dataprocessing workflow for the evaluation of the whole genome sequencing


Error rates in nanopore-sequencing

A drawback to nanopore-sequencing is that it tends to be quite error prone. The error rates in nanopore sequencing can be as high as 15% [3], as opposed to, for example, Next-generation Illumina sequencing, which has an average error rate of of 0.24% ± 0.06% [4]. It is possible to recognize and eliminate such mistakes and the quality will improve the longer the machine runs.


Results


for the tumor sample we optained we performed whole genome sequencing and amplicon sequencing to detect possible deletions and point mutaions

Figure 1: Coverage plots of a nanopore-sequencing run of tumor tissue at different time points. The upper plot shows the coverage after six hours, the plot below shows the coverage after one day
Median coverage of the human chromosomes. Upper graph displays the coverage after 6 h, the bottom graph displays the coverage after 24 h. Orange dots denote the median calculated for individual 500 kbp windows, blue dots denote the average of 10 adjacent windows.
We are able to detect the characteristic tert point mutation, which is neccesary for the characterisation of the glioma tumors, so we are able to identify potential targets for our mRNA based therapy. Based on this characterisation, we are able to use our sensor design tool and can push personalized into a new direction.


Our whole genome sequencing data showed areas indicating the presence of deletions (coverage significantly lower than the average) in several regions of the chromosomes. While mapping artefacts caused by mapping to repetitive elements need to be taken into account, the sequencing allows easy identification of indel events. However, we did not detect typical 1P and 19Q deletions that would have allowed a classification as oligodendroglioma. Based on the data generated in 24 h, we learned that using a single GridION flowcell would not suffice to generate enough data within the 3 h intraoperative window. This could be alliavated by either using several flowcells (GridION can run up to 5 in parallel) or by using a PromethION sequencer, as one PromethION flowcell genrates as much data as 6 GridION flowcells.


  • 1. Gayet, R. V. et al. Autocatalytic base editing for RNA-responsive translational control. Nat Commun 14, 1339 (2023).