| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: Mapping Cancer Markers Forum
Thread: MCM Progress Reports |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 957
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
For the around the past 100 MCM sequences they have been that tiny. They started in the MCM1_00043xx series and have continued.
|
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7244 Status: Offline Project Badges: 
|
Hello All, The research scientists are working on their next project update now and we should have it shortly. Thanks. Tedi Thank you Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
For the around the past 100 MCM sequences they have been that tiny. They started in the MCM1_00043xx series and have continued. Thanks for bringing that up. Suspended the network and looked at 20 in the transfer tab through batch 4429, then the actual file sizes of 148 of 150 byte. The project specifies output from 0.1mb to 3mb. Wonder what mapping data can be contained in so few bytes. Did you do a bit for bit file comparison of two of exact same size? Just curious if it's something like "nothing there". |
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3265 Status: Offline Project Badges: 
|
Thanks for the update.
----------------------------------------AMD Ryzen 5 1600AF 4C/8T 3.2 GHz - 85W AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W Intel Z3740 4C/4T 1.8 GHz - 6W |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I haven't compared the files. We have gone from seeing upwards of 2MB to seeing almost nothing. It is a possibility that these WU's just have no "hope" in terms of fighting cancer. It does seem weird though that we have 100 series all looking the same way. So either the lung WU run has been completed (earlier than expected) and we are now into the next group and this is normal. Another possibility, the scientists were working on a way to reduce the file sizes and something has gone awry in the process. I have no idea what pertinent information could be had in 150 bytes or under. Even with the best compression in the world, it is still a very small amount.
----------------------------------------Hopefully someone with some knowledge will come along with some information on if this is expected behavior. A low of .1MB would be in the range we are seeing. [Edit 1 times, last edit by Former Member at May 21, 2014 2:47:39 PM] |
||
|
|
cubes
World Community Grid Tech, Mapping Cancer Markers and Help Conquer Scientist Canada Joined: Mar 3, 2007 Post Count: 58 Status: Offline Project Badges: 
|
MCM is still grinding through lung cancer data. When we finish with lung cancer and switch to a new cancer dataset, we'll be sure to announce it!
lanbrown, the recent change you noticed is caused by a combination of two factors, a change in the search parameters, and greater (or shall we say overzealous?) efforts to control excessive result sizes. Until about two weeks ago, the lung cancer analysis had focused on gene signatures composed of 10 to 20 biomarkers, with some work units being random surveys of "signature space", and others being optimizing searches. The newest work units are looking at very short gene signatures (5 to 9 biomarkers) and longer ones (21-25 biomarkers), and are configured for random surveys. The time it takes to evaluate a single gene signature depends in part on the size of the signature. (As you might expect, larger signatures generally take more time to evaluate.) Signature size can also affect the work unit result size. As you've probably noticed, both the run-times and the result sizes vary in this project. Neither is perfectly predictable, though we are working to improve the consistency of the run time and to limit excessive result sizes. To control result sizes, we filter-out the lowest-performing signatures from the returned results. Each work unit specifies a minimum quality threshold that all returned results must exceed. For the first few batches with these new shorter and longer signatures, we chose a conservative threshold. We will adjust the filter in the near future. It's important to note that even though the current work units are not finding many successful gene signatures, these negative results are contributing to the statistical picture that this project is slowly building. Each work unit reports back how many signatures were tested, how many failed the test, plus details on the ones that passed. So every result contributes. But sadly, not every work unit finds a one-in-a-million signature. Christian A. Cumbaa Research Associate Princess Margaret Cancer Centre University Health Network |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
  Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
A well worded report that really makes me feel I know what's going on. Thank-you, Christian!
|
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2955 Status: Offline Project Badges: 
|
Yes, thank you Christian for your post - it's certainly good to hear from the researchers every now and again
----------------------------------------  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7244 Status: Offline Project Badges:
|
Thank you.
----------------------------------------
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
|