<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0"><channel><title>Disqus - Latest Comments for dalloliogm</title><link>http://disqus.com/by/dalloliogm/</link><description></description><atom:link href="http://disqus.com/dalloliogm/comments.rss" rel="self"></atom:link><language>en</language><lastBuildDate>Thu, 02 Nov 2023 08:39:00 -0000</lastBuildDate><item><title>Re: I rischi legati all’Intelligenza artificiale</title><link>https://opinione.it/hi-tech/2023/10/13/mario-sammarone_intelligenza-artificiale-bostrom-lungotermismo-musk-altman-cina/#comment-6314285392</link><description>&lt;p&gt;Hai proprio ragione, purtroppo un blocco allo sviluppo dell'AI sarebbe difficile da realizzare, e otterrebbe pochi risultati. Imprese come Google, Microsoft, Facebook spendono milioni di dollari già solo per fare il training di questi modelli, e difficilmente sarebbero contenti di abbandonare lo sviluppo. Anche se lo facessero, altre imprese in altri Paesi prenderebbero il posto. È difficile fermare la tecnologia, conviene cercare di approfittare della situazione e capite come utilizzare questi nuovi strumenti al meglio.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 02 Nov 2023 08:39:00 -0000</pubDate></item><item><title>Re: A Compelling Case For SparkR A Data Science Engineer’s thoughts on Applied Data Science</title><link>https://cosminsanda.com/posts/a-compelling-case-for-sparkr/#comment-5056234470</link><description>&lt;p&gt;Thank you for the article! For some reasons I thought SparkR was no longer being developed, but this does not seem to be true. The code examples you have posted are also very useful.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 03 Sep 2020 05:29:28 -0000</pubDate></item><item><title>Re: A practical guide to methods controlling false discoveries in computational biology</title><link>https://www.biorxiv.org/content/early/2018/10/31/458786#comment-4205473728</link><description>&lt;p&gt;Some comments about the paper:&lt;/p&gt;&lt;p&gt;- Since there are many colors used in Figure 2 and 3, it is difficult to read which method is which (especially if you printed the paper in grey scale, like I did :-) ).&lt;/p&gt;&lt;p&gt;- in Fig 4, it would be useful to have the maximum number of total rejection for each column. It took me a while to understand that the columns have different scale. For example, I would expect BH and q-value to have the same color intensity within each dataset (e.g. ChIPseq, etc..) as these methods do not make use of covariates. However the colors are different, even when the total number of tests is the same (e.g. in many scRNA datasets).&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Wed, 21 Nov 2018 04:57:49 -0000</pubDate></item><item><title>Re: A practical guide to methods controlling false discoveries in computational biology</title><link>https://www.biorxiv.org/content/early/2018/10/31/458786#comment-4203911753</link><description>&lt;p&gt;Thank you very much for the quick answer. So in a GWAS, one may think that allele frequency is not a good covariate, because se and p-values are calculated from it. However, under the null hypothesis allele frequency is independent on the test, so we can use it as a covariate.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 20 Nov 2018 06:55:04 -0000</pubDate></item><item><title>Re: A direct approach to estimating false discovery rates conditional on covariates</title><link>https://www.biorxiv.org/content/early/2018/01/22/035675#comment-4194714734</link><description>&lt;p&gt;This is an interesting method, however I don't understand the application to GWASes. Isn't the p-value in a GWAS dependent on the allele frequency?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Wed, 14 Nov 2018 10:56:42 -0000</pubDate></item><item><title>Re: A practical guide to methods controlling false discoveries in computational biology</title><link>https://www.biorxiv.org/content/early/2018/10/31/458786#comment-4180826942</link><description>&lt;p&gt;The methods presented in the paper are very interesting. However there is still something I don't fully understand yet. The authors suggest to use independent covariates when adjusting for FDR. For example, in a eQTL study you may want to give more importance to the SNPs closer to the target gene, as these are more likely to be real associations. However, what is the advantage of including these weights at the FDR level, instead of covariates during the previous analysis? Shouldn't a FDR correction be agnostic to any hypothesis?&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 06 Nov 2018 04:33:43 -0000</pubDate></item><item><title>Re: Addressing the digital divide in contemporary biology: Lessons from teaching UNIX</title><link>http://www.biorxiv.org/content/early/2017/04/07/122424#comment-3536846051</link><description>&lt;p&gt;Very good materials!&lt;/p&gt;&lt;p&gt;On the section "Why should we use Unix" I would insist more on explaining the Unix Approach. Even if outdated, the approach of writing "One program that does one thing and does it well", and to write tools that can be combined together, is useful for people learning programming in general, even if they don't end up using Unix daily.&lt;/p&gt;&lt;p&gt;Second, I publicly challenge the authors to include more "cows" in their exercises, following some materials I wrote for another course: &lt;a href="https://nbviewer.jupyter.org/format/slides/github/dalloliogm/belgrade_unix_intro/blob/master/PEB%20Bash%20Workshop.ipynb#/" rel="nofollow noopener" target="_blank" title="https://nbviewer.jupyter.org/format/slides/github/dalloliogm/belgrade_unix_intro/blob/master/PEB%20Bash%20Workshop.ipynb#/"&gt;https://nbviewer.jupyter.or...&lt;/a&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 26 Sep 2017 12:28:34 -0000</pubDate></item><item><title>Re: “Caro Mattarella, intervenga contro i test di Medicina”</title><link>http://www.ilfattoquotidiano.it/premium/articoli/caro-mattarella-intervenga-contro-i-test-di-medicina/#comment-2848226961</link><description>&lt;p&gt;Non sono d'accordo, io il test l'avevo superato senza seguire nessun corso. Basta mettersi a studiare.. &lt;br&gt;&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Sat, 20 Aug 2016 09:43:53 -0000</pubDate></item><item><title>Re: KEGG enrichment analysis with latest online data using clusterProfiler</title><link>https://guangchuangyu.github.io/2015/02/kegg-enrichment-analysis-with-latest-online-data-using-clusterprofiler/#comment-2561544724</link><description>&lt;p&gt;Yes, the readLines command seems to be working correctly.&lt;br&gt;I will try with KEGGREST. Do you know if there is a way  to use a locally downloaded copy of the KEGG database? We have the licence and access to the FTP, so I wonder if I can use that data instead.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 10 Mar 2016 06:56:01 -0000</pubDate></item><item><title>Re: KEGG enrichment analysis with latest online data using clusterProfiler</title><link>https://guangchuangyu.github.io/2015/02/kegg-enrichment-analysis-with-latest-online-data-using-clusterprofiler/#comment-2561443364</link><description>&lt;p&gt;Thanks!&lt;br&gt;I am working behind a HTTP proxy with password. What is the best way to overcome it?&lt;/p&gt;&lt;p&gt;I am getting this error:&lt;/p&gt;&lt;p&gt;&amp;gt; yy = enrichKEGG(gcSample[[5]])&lt;br&gt;Error in .getUrl(url, .listParser, nameColumn = 1, valueColumn = 2) :&lt;/p&gt;&lt;p&gt;  Proxy Authentication Required (HTTP 407).&lt;/p&gt;&lt;p&gt;Note that internet access works properly. I can download files with download.file and install new packages from bioconductor and cran.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 10 Mar 2016 05:01:56 -0000</pubDate></item><item><title>Re: Bringing the powers of SQL into R</title><link>https://datascienceplus.com/bringing-the-powers-of-sql-into-r/#comment-2409912108</link><description>&lt;p&gt;Nice article, but you should have a look at dplyr. It makes integration with databases much easier.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Mon, 14 Dec 2015 17:58:38 -0000</pubDate></item><item><title>Re: Genetics and racism (1) | The Grand Locus</title><link>http://blog.thegrandlocus.com/2012/12/genetics-and-racism-1#comment-1957052734</link><description>&lt;p&gt;Very nice post!&lt;br&gt;The early times of population genetics were actually quite racists, with PCA being used for phrenology and for distinguishing idiots/immigrants from the rest of people, and with the whole discipline being called eugenetics. &lt;br&gt;A nice book on the topic is "Fatal Invention" by Dorothy Roberts. and also Gould's The Mismeasure of Man. I used to work in the PRBB in Barcelona until two years ago, and before leaving I left a copy of Roberts' book in the shelves of the cafeteria. If it is still there, I recommend you to read it ;-)&lt;br&gt;I would be able to speak for hours about genetics and racism, because I think it is an interesting subject. When I wrote my thesis, I dedicated the preface to Blumenbach, the scientist who introduce the term "caucasian" and that involuntarily favored a racist approach in genetics afterwards (&lt;a href="http://bioinfoblog.it/2013/08/thesis-deposited-here-is-my-preface/" rel="nofollow noopener" target="_blank" title="http://bioinfoblog.it/2013/08/thesis-deposited-here-is-my-preface/"&gt;http://bioinfoblog.it/2013/...&lt;/a&gt; ). &lt;br&gt;Thanks for your post!&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Fri, 10 Apr 2015 07:51:49 -0000</pubDate></item><item><title>Re: Partiti, nel 2013 hanno speso 45 milioni. Ma se ne sono fatti rimborsare 54</title><link>http://www.ilfattoquotidiano.it/2015/03/27/partiti-nel-2013-speso-45-milioni-se-ne-fatti-rimborsare-54/1543703/#comment-1931718387</link><description>&lt;p&gt;Con 8 milioni di euro si sarebbe potuto salvare il Mario Negri sud in Abruzzo. Invece si è preferito mandare tutti i ricercatori a casa, senza arretrati né tfr.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Fri, 27 Mar 2015 14:01:13 -0000</pubDate></item><item><title>Re: Internazionale » Opinioni » A lezione di storia</title><link>http://www.internazionale.it/opinioni/giulia-zoli/2014/07/29/a-lezione-di-storia/#comment-1510180321</link><description>&lt;p&gt;L'Etiopia è stato l'ultimo paese africano ad essere colonizzato da una potenza europea, tanto che all'inizio del secolo si utilizzava la parola Etiopia per riferirsi a tutta l'Africa, un poco come quando Taiwan era sinonimo di Cina.&lt;br&gt;La colonizzazione italiana in realtà fu dovuta più all'azione politica che alla conquista militare. I nostri eserciti erano molto male organizzati, e sul campo di battaglia la guerra di conquista fu persa. Solo l'intervento delle altre potenze europee, per un gioco di equilibrio politico,  sancì l'annessione dell'Etiopia all'Italia.&lt;br&gt;Tra l'altro la guerra di liberazione dell'Etiopia dal dominio italiano, e dalla supremazia europea (che sancì l'annessione nonostante la guerra fosse stata persa), è alla base della religione rastafarianesima. La prossima volta che vedi una persona con capelli rasta, sappi che in realtà sta protestando contro l'occupazione italiana in Etiopia - o più in generale, per la libertà dei popoli.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 29 Jul 2014 11:01:02 -0000</pubDate></item><item><title>Re: Sperimentazione animale, Lav: “E&amp;#8217; un errore”. Scienziati: “Indispensabile”</title><link>http://www.ilfattoquotidiano.it/2013/09/26/sperimentazione-animale-lav-nessun-compromesso-scienziati-in-allarme/699619/#comment-1059751733</link><description>&lt;p&gt;servirebbero milioni di volontari...&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 26 Sep 2013 09:19:11 -0000</pubDate></item><item><title>Re: Sperimentazione animale, Lav: “E&amp;#8217; un errore”. Scienziati: “Indispensabile”</title><link>http://www.ilfattoquotidiano.it/2013/09/26/sperimentazione-animale-lav-nessun-compromesso-scienziati-in-allarme/699619/#comment-1059747820</link><description>&lt;p&gt;Non capisco la ragione di tanta ostilitá verso scienziati e professori. Si tratta di persone che lavorano guadagnando pochi soldi e facendo molti sacrifici.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 26 Sep 2013 09:14:59 -0000</pubDate></item><item><title>Re: Sperimentazione animale, Lav: “E&amp;#8217; un errore”. Scienziati: “Indispensabile”</title><link>http://www.ilfattoquotidiano.it/2013/09/26/sperimentazione-animale-lav-nessun-compromesso-scienziati-in-allarme/699619/#comment-1059745002</link><description>&lt;p&gt;In parte, i volontari umani si usano giá. Ogni farmaco, per essere approvato dalla FDA, deve passare diverse fasi di sperimentazione. Le prime fasi sono fatte su animali; le ultime sono fatte su umani volontari.&lt;/p&gt;&lt;p&gt;Purtroppo sarebbe impossibile trovare tanti volontari umani per sostituire la fase di sperimentazione su animali. Quante persone sarebbero disposte a farsi iniettare nuove sostanze, delle cuali non si sa niente e che non sono state mai provate su nessun organismo? Quante persone sarebbero disposte a farsi trapiantare organi e a subire operazioni chirurgiche, senza che queste siano state provate prima su altri organismi? Anche pagando una cifra considerevole, penso che sarebbe proprio difficile trovare tanti volontari.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 26 Sep 2013 09:13:06 -0000</pubDate></item><item><title>Re: Sperimentazione animale, Lav: “E&amp;#8217; un errore”. Scienziati: “Indispensabile”</title><link>http://www.ilfattoquotidiano.it/2013/09/26/sperimentazione-animale-lav-nessun-compromesso-scienziati-in-allarme/699619/#comment-1059671442</link><description>&lt;p&gt;Io, in qualitá di ricercatore e amante della natura, sono addolorato per la sperimentazione animale, e capisco le motivazioni della LAV. Purtroppo peró non c'è nessuna reale alternativa. Nessun modello computazionale o altro sistema sará mai una valido sostituto alla sperimentazione. Non è possibile pensare di saltare i test su animali prima di avviare la sperimentazione su umani.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 26 Sep 2013 07:54:59 -0000</pubDate></item><item><title>Re: Elezioni 2013, Giannino: &amp;#8220;Pronto al passo indietro, deciderà la direzione nazionale&amp;#8221;</title><link>http://www.ilfattoquotidiano.it/2013/02/19/giannino-pronto-al-passo-indietro-decidera-direzione-nazionale/505281/#comment-804338051</link><description>&lt;p&gt;Bravo, vediamo se riesci a invalidare il voto degli italiani all'estero, che giá abbiamo votato...&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 19 Feb 2013 09:53:13 -0000</pubDate></item><item><title>Re: Anecdotal science</title><link>http://ivory.idyll.org/blog/anecdotal-science.html#comment-641995243</link><description>&lt;p&gt;It must also be said that the way most current software for version control is designed is not optimal for bioinformatics. They are excellent for developing applications or websites; but when it comes to research, they have some limits.&lt;/p&gt;&lt;p&gt;First, in bioinformatics it is much more frequent to reutilize some of the scripts among different projects. For example, I can have a script to parse FASTA files, which I would use in different separate projects. What is the best way to source control this script? One solution is to copy the script from one project to another, including a reference to the original repository and version in the commit description. One other solution is to transform the script into a library... but it takes more time, and it may lead you to an hell of smaller repositories and dependencies.&lt;/p&gt;&lt;p&gt;Second, most software for version control can't handle big datasets efficiently. Mercurial has some extensions to keep track of big files, but it's not enough. Try to keep track of a few Gigabytes of data... You will see that these software are not designed for this use case.&lt;/p&gt;&lt;p&gt;Third, most bioinformatics are trapped between two extremes: the group leader who usually is not much interested in the programming details, and the need to present him good results. Your boss will probably not understand what software control is, and he will see it as something that makes you waste time. It is really much difficult to convince somebody that has a lot more research experience than you, that you need time to keep version control of your code. The fruits of good practices usually arrive only when the work is done, and sometimes that comes too late if there is somebody pressing you.&lt;/p&gt;&lt;p&gt;This being said, I am a strong supporter of using version control for bioinformatics. All the code from my last paper is available on bitbucket (&lt;a href="https://bitbucket.org/dalloliogm/pathways" rel="nofollow noopener" target="_blank" title="https://bitbucket.org/dalloliogm/pathways"&gt;https://bitbucket.org/dallo...&lt;/a&gt; ) and the future papers will be documented much better (meaning that I am using it for all my current research projects, but I can't show them as long as they are embargoed).&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Thu, 06 Sep 2012 10:30:52 -0000</pubDate></item><item><title>Re: felipe intervista felipe (dettagli sulla chiusura)</title><link>http://pollycoke.net/2009/09/09/felipe-intervista-felipe-dettagli-sulla-chiusura/#comment-16498439</link><description>&lt;p&gt;Mi va bene che questo blog abbia chiuso, solo i migliori sono capaci di decidere quando é il momento di smettere e di lanciarsi in un nuovo campo :-).&lt;/p&gt;&lt;p&gt;Ciononostante mi mancherà pollycoke, é stata una lettura piacevole per alcuni anni e mi ha insegnato tante cose e buoni trucchi su Linux e Gnome.&lt;/p&gt;&lt;p&gt;L'unico commento negativo che ho, é che avresti potuto lasciare lo scettro a qualcun altro, trovare gente disponibile a scrivere su questo blog o almeno indicare i blog di tuo gradimento, quelli che consiglieresti agli 'orfani di pollycoke' :-)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Sat, 12 Sep 2009 11:07:34 -0000</pubDate></item><item><title>Re: Chi ha paura di Google Chrome OS?</title><link>http://pollycoke.net/2009/07/13/chi-ha-paura-di-google-chrome-os/#comment-12576176</link><description>&lt;p&gt;non lo useró, rimarró con Linux (debian o ubuntu)&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Mon, 13 Jul 2009 10:00:35 -0000</pubDate></item><item><title>Re: Linux sui Netbook: i produttori hanno paura di Microsoft</title><link>http://pollycoke.net/2009/06/20/linux-sui-netbook-i-produttori-hanno-paura-di-microsoft/#comment-11501623</link><description>&lt;p&gt;Non credo che le cose stiano così... Molte persone adesso sono abituate a Windows, e dopo aver provato Linux su netbook si sono trovate male non solo per l'interfaccia diversa, ma anche per la mancanza di molti programmi famigliari, la parziale incompatibilità dei formati, e la mancanza di supporto tecnico valido da parte dei rivenditori di pc.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Sat, 20 Jun 2009 11:58:47 -0000</pubDate></item><item><title>Re: Rubano anche il vecchio logo di Arch Linux</title><link>http://pollycoke.net/2009/06/08/rubano-anche-il-vecchio-logo-di-arch-linux/#comment-10646673</link><description>&lt;p&gt;va bene, ma é chiaro che si tratta solo di un caso. Il proprietario di quella compagnia probabilmente non aveva nemmeno idea di cosa fosse Linux, tantomeno di arch, e sicuramente denunciarlo non é la miglior maniera per farglielo conoscere.&lt;/p&gt;&lt;p&gt;Se si fosse trattato di una societá di software o di consulenza su windows sarebbe potuto essere sgradevole, ma questo é vittimismo...&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Tue, 09 Jun 2009 06:40:17 -0000</pubDate></item><item><title>Re: Rubano anche il vecchio logo di Arch Linux</title><link>http://pollycoke.net/2009/06/08/rubano-anche-il-vecchio-logo-di-arch-linux/#comment-10624837</link><description>&lt;p&gt;ma dai é solo una stupida coincidenza... vivi e lascia vivere, a chi vuoi che gliene importi che il logo di una distribuzione linux e di una compagnia siano uguali?&lt;/p&gt;&lt;p&gt;Questo é comportarsi come la Free Software Fundation, é perdere tempo e dare una impressione negativa su Linux.&lt;/p&gt;</description><dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">dalloliogm</dc:creator><pubDate>Mon, 08 Jun 2009 16:01:13 -0000</pubDate></item></channel></rss>