What specifications of laptop should I buy for bioinformatics?

Question

[yonghyun09]

Hello, I have a question regarding computer settings.

The lab I belong to is currently in the process of setting up microbiome research, and is experiencing difficulties in purchasing a computer needed for BI analysis.

Qiime2 analysis was performed by installing a virtual machine on a Windows computer with 16GB of RAM, (in VirtualBox, 10GB) but the use of DADA2, Classifier was limited due to lack of performance, and that steps were able to be taken with the cooperation of other research institutes.

I am currently planning to purchase a computer with suitable specifications (planning to set up a mac or Linux system), and I would like to ask if I can get any recommendations or specifications.

I am also considering purchasing a computer for personal investment(for BI analysis), so I am also considering a laptop such as below. (The cost is under consideration at around $ 2,000 or less.)

• Macbook Pro 13, M2, RAM 24GB
• DELL Inspiron 16 DN5620-UB05KR, i7, RAM 32GB
• DELL Inspiron 14 DN5420-UB02KR, i5, RAM 32GB
• HP Elitebook 650 G9-4W5J8AV, i5, RAM 64GB

However, Mac's M chip has been pointed out about the disadvantages of long-term use due to compatibility issues compared to intel, and I would like to ask for advice on this. Which laptop is best to buy and use?

And as considered above, as a personal computer, I'm considering using a laptop because the desktop would be inconvenient. Our laboratory does not have a server, so analysis using a server is expected to be limited.

I would be very grateful if you could give me some advice on these difficulties.

Thank you.

Answer 1

[lacb]

Memory can be a limiting factor on many pipelines with some processes requiring up to 50Go of RAM. I would suggest testing your pipeline on a computer with a lot of RAM and CPU to have an idea of what you need for your pipeline before buying your laptop, this also depends on the size of your input files, how many parallel jobs you want to run, etc.

Also you can check the CPU for different instruction sets (AVX2, SSE2, etc.) because they can speed up a lot compatible programs. Generally, recent Intel CPUs have good standards for scientific computing.

[yonghyun09]

Thank you for your kind reply. As you said, it is appropriate to test the product in advance with a computer with various specifications before making a purchase after understanding the requirements. However, considering my current environment, there is no opportunity to do such a test in advance. Therefore, if you are in a situation where you have to make a purchase with insufficient information, may I ask which laptop you would choose from the above?

One piece of information that can be referenced is that the analysis, which did not run on my existing Windows desktop virtual machine (RAM 10GB), was completed by running for about 8 hours on another institution's imac (RAM 8GB).

[lacb]

I think I would discard the Macbook because I don't know a lot with scientific computing on Mac.

Regarding the two DELL laptop, I would look like the two CPU to check their extensions (AVX/SSE/...), if they are the same then the choice depends on the computing power/speed you need (RAM is the same).

The HP/64 GB has more RAM, I think this can be very useful because RAM could be a limiting factor, making your pipeline to fail if there is not enough. Whereas a worse CPU would "just" slow down your analysis. If there is much RAM available and CPUs are underused, you can also consider parallelizing your pipeline, but this depends on the CPU/RAM requirement and the degree of parallelization of your pipeline. Also, RAM allows caching more temporary files and sometimes accelerates the pipeline when I/O is limiting.

I think I'd choose the HP/64GB, but I'm highly biased by my own usage, so I'd recommend finding out more on your specific pipeline requirements.

[yonghyun09]

Thank you Your answer was very helpful! I will consider the needs of RAM and CPU carefully before purchasing.

Answer 2

[Mensur Dlakic]

And as considered above, as a personal computer, I'm considering using a laptop because the desktop would be inconvenient. Our laboratory does not have a server, so analysis using a server is expected to be limited.

Since you want any advice, mine would be to buy a desktop with a lot more memory than 32-64 Gb. In my experience, that memory size is not enough for most serious bioinformatics applications. I am finding out more recently that even 256 Gb often doesn't work. On top of that, laptops typically come with 0.5-1 Tb disk (or even 0.25 Tb), which is not easy to extend and it isn't enough for serious datasets. Desktops will come with at least 2-4 Tb disks, and it is usually easy enough to add a couple of 10 Tb puppies to them because they have more space for extra drives.

If you still don't want to / can't buy a desktop, I think at the very least you should secure access to some kind of high performance computing (HPC) cluster. Then you can do heavy lifting on a cluster, and you could get away with a 32 Gb laptop for supporting processes and simple data analysis. If you have never experienced a sound of a laptop fan that is doing heavy-duty calculation, trust me that it isn't pretty. Top priority in laptops is mobility (light, long battery life), not ability to do heavy work.

[yonghyun09]

Thank you for your kind and detailed reply. As you said, bioinformatics technology is developing rapidly, and it seems that the required computer specifications are increasing accordingly. It seems to be important to build conditions that can use high desktop specifications and HPC clusters. However, since the lab I belong to is in a small and low-budget environment and in the early stages of NGS-based analysis, it seems that building a good computer right away will be very limited.

I am conducting a microbiome modulation study using bacteriophage, but there is no need for a very large-scale NGS analysis yet. Therefore, i will carefully refer to the advice you gave and purchase the computer that can be secured as much as possible. Thanks again for your kind reply.